From f2f12b6fc032c7b1419fd6db84e2868b5f05a878 Mon Sep 17 00:00:00 2001 From: Shuah Khan Date: Wed, 6 Jun 2012 10:50:06 -0600 Subject: iommu/amd: Fix missing iommu_shutdown initialization in passthrough mode The iommu_shutdown callback is not initialized when the AMD IOMMU driver runs in passthrough mode. Fix that by moving the callback initialization before the check for passthrough mode. Signed-off-by: Shuah Khan Cc: stable@vger.kernel.org Signed-off-by: Joerg Roedel diff --git a/drivers/iommu/amd_iommu_init.c b/drivers/iommu/amd_iommu_init.c index 542024b..c04ddca 100644 --- a/drivers/iommu/amd_iommu_init.c +++ b/drivers/iommu/amd_iommu_init.c @@ -1641,6 +1641,8 @@ static int __init amd_iommu_init(void) amd_iommu_init_api(); + x86_platform.iommu_shutdown = disable_iommus; + if (iommu_pass_through) goto out; @@ -1649,8 +1651,6 @@ static int __init amd_iommu_init(void) else printk(KERN_INFO "AMD-Vi: Lazy IO/TLB flushing enabled\n"); - x86_platform.iommu_shutdown = disable_iommus; - out: return ret; -- cgit v0.10.2 From ea3b2ea24ef0f2ef9c6795b19cff456195b6728a Mon Sep 17 00:00:00 2001 From: Shmulik Ladkani Date: Fri, 8 Jun 2012 18:29:06 +0300 Subject: mtd: nand: initialize bitflip_threshold prior to BBT scanning As of edbc454 [mtd: driver _read() returns max_bitflips; mtd_read() returns -EUCLEAN], 'mtd->bitflip_threshold' must be set for mtd devices having ECC, prior any 'mtd_read()' call. Otherwise, 'mtd_read()' will falsely return -EUCLEAN. Normally, 'mtd->bitflip_threshold' is initialized when the MTD is added. However, this is too late for NAND MTDs, as 'scan_bbt()' is invoked prior the existing initialization of 'mtd->bitflip_threshold'. This is a problem since 'scan_bbt()' calls 'mtd_read()', in the case of a flash-based bad block table. It resulted in a falsely reported bitflips indication during BBT read, which lead to constant scrubbing of the flash BBT blocks. Initialize 'mtd->bitflip_threshold' to its default value (if not already set by the driver), prior to invocation of 'scan_bbt()'. Reported-by: Sascha Hauer Tested-by: Sascha Hauer Signed-off-by: Shmulik Ladkani Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c index d47586c..a11253a0 100644 --- a/drivers/mtd/nand/nand_base.c +++ b/drivers/mtd/nand/nand_base.c @@ -3501,6 +3501,13 @@ int nand_scan_tail(struct mtd_info *mtd) /* propagate ecc info to mtd_info */ mtd->ecclayout = chip->ecc.layout; mtd->ecc_strength = chip->ecc.strength; + /* + * Initialize bitflip_threshold to its default prior scan_bbt() call. + * scan_bbt() might invoke mtd_read(), thus bitflip_threshold must be + * properly set. + */ + if (!mtd->bitflip_threshold) + mtd->bitflip_threshold = mtd->ecc_strength; /* Check, if we should skip the bad block table scan */ if (chip->options & NAND_SKIP_BBTSCAN) -- cgit v0.10.2 From 2177905ca7419c49910d47e38e44790affd918cc Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Mon, 11 Jun 2012 23:56:09 -0700 Subject: Input: fix input.h kernel-doc warning Fix kernel-doc warning in input.h: Warning(include/linux/input.h:140): No description found for parameter 'len' Signed-off-by: Randy Dunlap Signed-off-by: Dmitry Torokhov diff --git a/include/linux/input.h b/include/linux/input.h index a816714..2740d08 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -116,6 +116,7 @@ struct input_keymap_entry { /** * EVIOCGMTSLOTS(len) - get MT slot values + * @len: size of the data buffer in bytes * * The ioctl buffer argument should be binary equivalent to * -- cgit v0.10.2 From e73f843a3235a19de38359c91586e9eadef12238 Mon Sep 17 00:00:00 2001 From: Suresh Jayaraman Date: Tue, 12 Jun 2012 07:15:50 +0530 Subject: cifs: fix parsing of password mount option The double delimiter check that allows a comma in the password parsing code is unconditional. We set "tmp_end" to the end of the string and we continue to check for double delimiter. In the case where the password doesn't contain a comma we end up setting tmp_end to NULL and eventually setting "options" to "end". This results in the premature termination of the options string and hence the values of UNCip and UNC are being set to NULL. This results in mount failure with "Connecting to DFS root not implemented yet" error. This error is usually not noticable as we have password as the last option in the superblock mountdata. But when we call expand_dfs_referral() from cifs_mount() and try to compose mount options for the submount, the resulting mountdata will be of the form ",ver=1,user=foo,pass=bar,ip=x.x.x.x,unc=\\server\share" and hence results in the above error. This bug has been seen with older NAS servers running Samba 3.0.24. Fix this by moving the double delimiter check inside the conditional loop. Changes since -v1 - removed the wrong strlen() micro optimization. Signed-off-by: Suresh Jayaraman Acked-by: Sachin Prabhu Cc: stable@vger.kernel.org [3.1+] Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 78db68a..5b38407 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1653,24 +1653,26 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, * If yes, we have encountered a double deliminator * reset the NULL character to the deliminator */ - if (tmp_end < end && tmp_end[1] == delim) + if (tmp_end < end && tmp_end[1] == delim) { tmp_end[0] = delim; - /* Keep iterating until we get to a single deliminator - * OR the end - */ - while ((tmp_end = strchr(tmp_end, delim)) != NULL && - (tmp_end[1] == delim)) { - tmp_end = (char *) &tmp_end[2]; - } + /* Keep iterating until we get to a single + * deliminator OR the end + */ + while ((tmp_end = strchr(tmp_end, delim)) + != NULL && (tmp_end[1] == delim)) { + tmp_end = (char *) &tmp_end[2]; + } - /* Reset var options to point to next element */ - if (tmp_end) { - tmp_end[0] = '\0'; - options = (char *) &tmp_end[1]; - } else - /* Reached the end of the mount option string */ - options = end; + /* Reset var options to point to next element */ + if (tmp_end) { + tmp_end[0] = '\0'; + options = (char *) &tmp_end[1]; + } else + /* Reached the end of the mount option + * string */ + options = end; + } /* Now build new password string */ temp_len = strlen(value); -- cgit v0.10.2 From e001f1c8e9c1d3b2f0c638bc712bcac10294aae3 Mon Sep 17 00:00:00 2001 From: Stephen Warren Date: Thu, 7 Jun 2012 17:57:36 -0600 Subject: of: export of_platform_populate() Without this, modules can't use this API, leading to build failures. Signed-off-by: Stephen Warren Signed-off-by: Rob Herring diff --git a/drivers/of/platform.c b/drivers/of/platform.c index 343ad29..3132ea0 100644 --- a/drivers/of/platform.c +++ b/drivers/of/platform.c @@ -462,4 +462,5 @@ int of_platform_populate(struct device_node *root, of_node_put(root); return rc; } +EXPORT_SYMBOL_GPL(of_platform_populate); #endif /* CONFIG_OF_ADDRESS */ -- cgit v0.10.2 From ae2a7d20087f2780df7ddc829a74f4a387f26b4d Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Thu, 12 Apr 2012 13:24:17 +0200 Subject: of: Add Avionic Design vendor prefix This commit adds a device tree vendor prefix for Avionic Design GmbH. Cc: Grant Likely Cc: Rob Herring Cc: devicetree-discuss@lists.ozlabs.org Cc: Linus Walleij Signed-off-by: Thierry Reding Signed-off-by: Rob Herring diff --git a/Documentation/devicetree/bindings/vendor-prefixes.txt b/Documentation/devicetree/bindings/vendor-prefixes.txt index 6eab917..db4d3af 100644 --- a/Documentation/devicetree/bindings/vendor-prefixes.txt +++ b/Documentation/devicetree/bindings/vendor-prefixes.txt @@ -3,6 +3,7 @@ Device tree binding vendor prefix registry. Keep list in alphabetical order. This isn't an exhaustive list, but you should add new prefixes to it before using them to avoid name-space collisions. +ad Avionic Design GmbH adi Analog Devices, Inc. amcc Applied Micro Circuits Corporation (APM, formally AMCC) apm Applied Micro Circuits Corporation (APM) -- cgit v0.10.2 From 37797d5f2ee47a06cc1c43b7b5f4832822436e3b Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Sat, 21 Apr 2012 14:21:13 -0300 Subject: dt: fsl-imx-cspi.txt: Fix comment about GPIOs used for chip selects The gpio numbers does not match the comment. Fix the comments for GPIOs used for chip selects. Signed-off-by: Fabio Estevam Signed-off-by: Rob Herring diff --git a/Documentation/devicetree/bindings/spi/fsl-imx-cspi.txt b/Documentation/devicetree/bindings/spi/fsl-imx-cspi.txt index 9841057..4256a6d 100644 --- a/Documentation/devicetree/bindings/spi/fsl-imx-cspi.txt +++ b/Documentation/devicetree/bindings/spi/fsl-imx-cspi.txt @@ -17,6 +17,6 @@ ecspi@70010000 { reg = <0x70010000 0x4000>; interrupts = <36>; fsl,spi-num-chipselects = <2>; - cs-gpios = <&gpio3 24 0>, /* GPIO4_24 */ - <&gpio3 25 0>; /* GPIO4_25 */ + cs-gpios = <&gpio3 24 0>, /* GPIO3_24 */ + <&gpio3 25 0>; /* GPIO3_25 */ }; -- cgit v0.10.2 From 06693fcd57739d2c6cb22c0f2bb494e7e65ccedd Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Sun, 10 Jun 2012 14:24:07 -0300 Subject: dt: fsl-imx-esdhc.txt: Fix gpio number assignment Fix gpio number assignment, so that it can match the number stated in the comment. Cc: Signed-off-by: Fabio Estevam Signed-off-by: Rob Herring diff --git a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt index c7e404b..fea541e 100644 --- a/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt +++ b/Documentation/devicetree/bindings/mmc/fsl-imx-esdhc.txt @@ -29,6 +29,6 @@ esdhc@70008000 { compatible = "fsl,imx51-esdhc"; reg = <0x70008000 0x4000>; interrupts = <2>; - cd-gpios = <&gpio0 6 0>; /* GPIO1_6 */ - wp-gpios = <&gpio0 5 0>; /* GPIO1_5 */ + cd-gpios = <&gpio1 6 0>; /* GPIO1_6 */ + wp-gpios = <&gpio1 5 0>; /* GPIO1_5 */ }; -- cgit v0.10.2 From d7858de94584a2edb842fa5bea1c5c7aa08608c4 Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Sun, 10 Jun 2012 14:24:08 -0300 Subject: dt: fsl-mma8450.txt: Add missing 'reg' description Add missing 'reg' description. Signed-off-by: Fabio Estevam ---- .../devicetree/bindings/input/fsl-mma8450.txt | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) Signed-off-by: Rob Herring diff --git a/Documentation/devicetree/bindings/input/fsl-mma8450.txt b/Documentation/devicetree/bindings/input/fsl-mma8450.txt index a00c94c..0b96e57 100644 --- a/Documentation/devicetree/bindings/input/fsl-mma8450.txt +++ b/Documentation/devicetree/bindings/input/fsl-mma8450.txt @@ -2,6 +2,7 @@ Required properties: - compatible : "fsl,mma8450". +- reg: the I2C address of MMA8450 Example: -- cgit v0.10.2 From af634df194955109197f0fe098d09eab741c0ce8 Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Sun, 10 Jun 2012 14:24:09 -0300 Subject: dt: fsl-fec.txt: Fix gpio number assignment Fix gpio number assignment, so that it can match the number stated in the comment. Signed-off-by: Fabio Estevam Signed-off-by: Rob Herring diff --git a/Documentation/devicetree/bindings/net/fsl-fec.txt b/Documentation/devicetree/bindings/net/fsl-fec.txt index 7ab9e1a..4616fc2 100644 --- a/Documentation/devicetree/bindings/net/fsl-fec.txt +++ b/Documentation/devicetree/bindings/net/fsl-fec.txt @@ -19,6 +19,6 @@ ethernet@83fec000 { reg = <0x83fec000 0x4000>; interrupts = <87>; phy-mode = "mii"; - phy-reset-gpios = <&gpio1 14 0>; /* GPIO2_14 */ + phy-reset-gpios = <&gpio2 14 0>; /* GPIO2_14 */ local-mac-address = [00 04 9F 01 1B B9]; }; -- cgit v0.10.2 From d44115f28edba06d0844ab4b79682f48500020f6 Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Sun, 10 Jun 2012 14:24:10 -0300 Subject: dt: mc13xxx.txt: Fix gpio number assignment Fix gpio number assignment, so that it can match the number stated in the comment. Signed-off-by: Fabio Estevam Signed-off-by: Rob Herring diff --git a/Documentation/devicetree/bindings/mfd/mc13xxx.txt b/Documentation/devicetree/bindings/mfd/mc13xxx.txt index 19f6af4..baf0798 100644 --- a/Documentation/devicetree/bindings/mfd/mc13xxx.txt +++ b/Documentation/devicetree/bindings/mfd/mc13xxx.txt @@ -46,8 +46,8 @@ Examples: ecspi@70010000 { /* ECSPI1 */ fsl,spi-num-chipselects = <2>; - cs-gpios = <&gpio3 24 0>, /* GPIO4_24 */ - <&gpio3 25 0>; /* GPIO4_25 */ + cs-gpios = <&gpio4 24 0>, /* GPIO4_24 */ + <&gpio4 25 0>; /* GPIO4_25 */ status = "okay"; pmic: mc13892@0 { -- cgit v0.10.2 From 107a84e61cdd3406c842a0e4be7efffd3a05dba6 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Thu, 14 Jun 2012 09:12:35 +0200 Subject: of: match by compatible property first When matching devices against an OF device ID table, the first string of the compatible property that is listed in the table should match, regardless of its position in the table. Cc: Grant Likely Cc: Rob Herring Cc: devicetree-discuss@lists.ozlabs.org Signed-off-by: Thierry Reding Signed-off-by: Rob Herring diff --git a/drivers/of/base.c b/drivers/of/base.c index d9bfd49..eada3f4 100644 --- a/drivers/of/base.c +++ b/drivers/of/base.c @@ -511,6 +511,22 @@ out: } EXPORT_SYMBOL(of_find_node_with_property); +static const struct of_device_id *of_match_compat(const struct of_device_id *matches, + const char *compat) +{ + while (matches->name[0] || matches->type[0] || matches->compatible[0]) { + const char *cp = matches->compatible; + int len = strlen(cp); + + if (len > 0 && of_compat_cmp(compat, cp, len) == 0) + return matches; + + matches++; + } + + return NULL; +} + /** * of_match_node - Tell if an device_node has a matching of_match structure * @matches: array of of device match structures to search in @@ -521,9 +537,18 @@ EXPORT_SYMBOL(of_find_node_with_property); const struct of_device_id *of_match_node(const struct of_device_id *matches, const struct device_node *node) { + struct property *prop; + const char *cp; + if (!matches) return NULL; + of_property_for_each_string(node, "compatible", prop, cp) { + const struct of_device_id *match = of_match_compat(matches, cp); + if (match) + return match; + } + while (matches->name[0] || matches->type[0] || matches->compatible[0]) { int match = 1; if (matches->name[0]) @@ -532,10 +557,7 @@ const struct of_device_id *of_match_node(const struct of_device_id *matches, if (matches->type[0]) match &= node->type && !strcmp(matches->type, node->type); - if (matches->compatible[0]) - match &= of_device_is_compatible(node, - matches->compatible); - if (match) + if (match && !matches->compatible[0]) return matches; matches++; } -- cgit v0.10.2 From 9e0304e388ef0e7495b279ce737c24ad56d20b5d Mon Sep 17 00:00:00 2001 From: Chris Metcalf Date: Mon, 18 Jun 2012 13:11:21 -0400 Subject: arch/tile: big-endian: properly bswap instruction bundles when backtracing Instruction bundles are always little-endian, even when running in big-endian mode. I missed this internal bug fix when cherry-picking the big-endian code to return to the community. Signed-off-by: Chris Metcalf diff --git a/arch/tile/kernel/backtrace.c b/arch/tile/kernel/backtrace.c index 9092ce8..f8b74ca 100644 --- a/arch/tile/kernel/backtrace.c +++ b/arch/tile/kernel/backtrace.c @@ -14,6 +14,7 @@ #include #include +#include #include #include #include @@ -336,8 +337,12 @@ static void find_caller_pc_and_caller_sp(CallerLocation *location, bytes_to_prefetch / sizeof(tile_bundle_bits); } - /* Decode the next bundle. */ - bundle.bits = prefetched_bundles[next_bundle++]; + /* + * Decode the next bundle. + * TILE always stores instruction bundles in little-endian + * mode, even when the chip is running in big-endian mode. + */ + bundle.bits = le64_to_cpu(prefetched_bundles[next_bundle++]); bundle.num_insns = parse_insn_tile(bundle.bits, pc, bundle.insns); num_info_ops = bt_get_info_ops(&bundle, info_operands); -- cgit v0.10.2 From 4d146ad7388a9cca2a890d7449aeb767565068d7 Mon Sep 17 00:00:00 2001 From: Fengguang Wu Date: Sat, 9 Jun 2012 20:00:19 -0300 Subject: [media] pms: fix build error in pms_probe() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix a compiler breakage introduced by commit 8173090acb33: drivers/media/video/pms.c: In function ‘pms_probe’: drivers/media/video/pms.c:1047:2: error: implicit declaration of function ‘kzalloc’ [-Werror=implicit-function-declaration] drivers/media/video/pms.c:1116:2: error: implicit declaration of function ‘kfree’ [-Werror=implicit-function-declaration] Signed-off-by: Fengguang Wu Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/pms.c b/drivers/media/video/pms.c index af2d908..77f9c921 100644 --- a/drivers/media/video/pms.c +++ b/drivers/media/video/pms.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include -- cgit v0.10.2 From ce7d16a1752accbff1ff7c813594aad0b7168563 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Thu, 21 Jun 2012 10:05:56 -0300 Subject: [media] smiapp-core: fix compilation build error smiapp-core.c:2472:3: error: implicit declaration of function 'kzalloc' [-Werror=implicit-function-declaration] Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/smiapp/smiapp-core.c b/drivers/media/video/smiapp/smiapp-core.c index e8c93c8..9cf5bda 100644 --- a/drivers/media/video/smiapp/smiapp-core.c +++ b/drivers/media/video/smiapp/smiapp-core.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include -- cgit v0.10.2 From b6a509df59c6abd0a2177e3be29642207711145d Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 20 Jun 2012 21:30:30 -0300 Subject: [media] media: pms.c needs linux/slab.h drivers/media/video/pms.c uses kzalloc() and kfree() so it should include to fix build errors and a warning. drivers/media/video/pms.c:1047:2: error: implicit declaration of function 'kzalloc' drivers/media/video/pms.c:1047:6: warning: assignment makes pointer from integer without a cast drivers/media/video/pms.c:1116:2: error: implicit declaration of function 'kfree' Found in mmotm but applies to mainline. Signed-off-by: Randy Dunlap Acked-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/pms.c b/drivers/media/video/pms.c index 77f9c921..b4c679b 100644 --- a/drivers/media/video/pms.c +++ b/drivers/media/video/pms.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include -- cgit v0.10.2 From ac1534a55d1e87d59a21c09c570605933b551480 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 21 Jun 2012 14:52:40 +0200 Subject: iommu/amd: Initialize dma_ops for hotplug and sriov devices When a device is added to the system at runtime the AMD IOMMU driver initializes the necessary data structures to handle translation for it. But it forgets to change the per-device dma_ops to point to the AMD IOMMU driver. So mapping actually never happens and all DMA accesses end in an IO_PAGE_FAULT. Fix this. Reported-by: Stefan Assmann Cc: stable@vger.kernel.org Signed-off-by: Joerg Roedel diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c index a2e418c..dfe7d37 100644 --- a/drivers/iommu/amd_iommu.c +++ b/drivers/iommu/amd_iommu.c @@ -83,6 +83,8 @@ static struct iommu_ops amd_iommu_ops; static ATOMIC_NOTIFIER_HEAD(ppr_notifier); int amd_iommu_max_glx_val = -1; +static struct dma_map_ops amd_iommu_dma_ops; + /* * general struct to manage commands send to an IOMMU */ @@ -2267,6 +2269,13 @@ static int device_change_notifier(struct notifier_block *nb, list_add_tail(&dma_domain->list, &iommu_pd_list); spin_unlock_irqrestore(&iommu_pd_list_lock, flags); + dev_data = get_dev_data(dev); + + if (!dev_data->passthrough) + dev->archdata.dma_ops = &amd_iommu_dma_ops; + else + dev->archdata.dma_ops = &nommu_dma_ops; + break; case BUS_NOTIFY_DEL_DEVICE: -- cgit v0.10.2 From 90e614bb4c581eb588ec26f130fcdd65aa047fa8 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Wed, 16 May 2012 11:35:04 -0300 Subject: [media] s5p-fimc: Fix bug in capture node open() When video pipeline initialization fails, the ST_CAPT_BUSY flag needs to be cleared before pm_runtime_put_sync is called. Otherwise the runtime suspend routine tries to suspend device, rather than just turning it off. Also fix potential null pointer dereference in fimc_pipeline_shutdown(). Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-capture.c b/drivers/media/video/s5p-fimc/fimc-capture.c index 3545745..7083107 100644 --- a/drivers/media/video/s5p-fimc/fimc-capture.c +++ b/drivers/media/video/s5p-fimc/fimc-capture.c @@ -499,10 +499,10 @@ static int fimc_capture_open(struct file *file) if (ret < 0) { dev_err(&fimc->pdev->dev, "Video pipeline initialization failed\n"); + clear_bit(ST_CAPT_BUSY, &fimc->state); pm_runtime_put_sync(&fimc->pdev->dev); fimc->vid_cap.refcnt--; v4l2_fh_release(file); - clear_bit(ST_CAPT_BUSY, &fimc->state); return ret; } ret = fimc_capture_ctrls_create(fimc); diff --git a/drivers/media/video/s5p-fimc/fimc-mdevice.c b/drivers/media/video/s5p-fimc/fimc-mdevice.c index 6753c45..7450dcd 100644 --- a/drivers/media/video/s5p-fimc/fimc-mdevice.c +++ b/drivers/media/video/s5p-fimc/fimc-mdevice.c @@ -193,9 +193,13 @@ int __fimc_pipeline_shutdown(struct fimc_pipeline *p) int fimc_pipeline_shutdown(struct fimc_pipeline *p) { - struct media_entity *me = &p->subdevs[IDX_SENSOR]->entity; + struct media_entity *me; int ret; + if (!p || !p->subdevs[IDX_SENSOR]) + return -EINVAL; + + me = &p->subdevs[IDX_SENSOR]->entity; mutex_lock(&me->parent->graph_mutex); ret = __fimc_pipeline_shutdown(p); mutex_unlock(&me->parent->graph_mutex); -- cgit v0.10.2 From d0da3c3565a1dd3cdfa374038d6ccae4b90fe142 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Wed, 16 May 2012 13:08:30 -0300 Subject: [media] s5p-fimc: Don't create multiple active links to same sink entity The driver is supposed to create active media link from sensor N (or its corresponding s5p-mipi-csis entity) to FIMC.N by default. Instead s5p-mipi-csis.N entity gets always connected by a default active link to FIMC.N, regardless of there are parallel bus sensor entities already connected to FIMC.N. Correct this. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-mdevice.c b/drivers/media/video/s5p-fimc/fimc-mdevice.c index 7450dcd..dffe4da 100644 --- a/drivers/media/video/s5p-fimc/fimc-mdevice.c +++ b/drivers/media/video/s5p-fimc/fimc-mdevice.c @@ -502,12 +502,12 @@ static void fimc_md_unregister_entities(struct fimc_md *fmd) * @source: the source entity to create links to all fimc entities from * @sensor: sensor subdev linked to FIMC[fimc_id] entity, may be null * @pad: the source entity pad index - * @fimc_id: index of the fimc device for which link should be enabled + * @link_mask: bitmask of the fimc devices for which link should be enabled */ static int __fimc_md_create_fimc_sink_links(struct fimc_md *fmd, struct media_entity *source, struct v4l2_subdev *sensor, - int pad, int fimc_id) + int pad, int link_mask) { struct fimc_sensor_info *s_info; struct media_entity *sink; @@ -524,7 +524,7 @@ static int __fimc_md_create_fimc_sink_links(struct fimc_md *fmd, if (!fmd->fimc[i]->variant->has_cam_if) continue; - flags = (i == fimc_id) ? MEDIA_LNK_FL_ENABLED : 0; + flags = ((1 << i) & link_mask) ? MEDIA_LNK_FL_ENABLED : 0; sink = &fmd->fimc[i]->vid_cap.subdev.entity; ret = media_entity_create_link(source, pad, sink, @@ -556,7 +556,10 @@ static int __fimc_md_create_fimc_sink_links(struct fimc_md *fmd, if (!fmd->fimc_lite[i]) continue; - flags = (i == fimc_id) ? MEDIA_LNK_FL_ENABLED : 0; + if (link_mask & (1 << (i + FIMC_MAX_DEVS))) + flags = MEDIA_LNK_FL_ENABLED; + else + flags = 0; sink = &fmd->fimc_lite[i]->subdev.entity; ret = media_entity_create_link(source, pad, sink, @@ -618,9 +621,8 @@ static int fimc_md_create_links(struct fimc_md *fmd) struct s5p_fimc_isp_info *pdata; struct fimc_sensor_info *s_info; struct media_entity *source, *sink; - int i, pad, fimc_id = 0; - int ret = 0; - u32 flags; + int i, pad, fimc_id = 0, ret = 0; + u32 flags, link_mask = 0; for (i = 0; i < fmd->num_sensors; i++) { if (fmd->sensor[i].subdev == NULL) @@ -672,19 +674,20 @@ static int fimc_md_create_links(struct fimc_md *fmd) if (source == NULL) continue; + link_mask = 1 << fimc_id++; ret = __fimc_md_create_fimc_sink_links(fmd, source, sensor, - pad, fimc_id++); + pad, link_mask); } - fimc_id = 0; for (i = 0; i < ARRAY_SIZE(fmd->csis); i++) { if (fmd->csis[i].sd == NULL) continue; source = &fmd->csis[i].sd->entity; pad = CSIS_PAD_SOURCE; + link_mask = 1 << fimc_id++; ret = __fimc_md_create_fimc_sink_links(fmd, source, NULL, - pad, fimc_id++); + pad, link_mask); } /* Create immutable links between each FIMC's subdev and video node */ -- cgit v0.10.2 From d547ab66e275e2f6bf703572e943018ef7c391c5 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Wed, 16 May 2012 15:00:26 -0300 Subject: [media] s5p-fimc: Honour sizeimage in VIDIOC_S_FMT Allow memory buffer size to be increased by means of struct v4l2_pix_plane_format::sizeimage at VIDIOC_S_FMT ioctl. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-capture.c b/drivers/media/video/s5p-fimc/fimc-capture.c index 7083107..13df723 100644 --- a/drivers/media/video/s5p-fimc/fimc-capture.c +++ b/drivers/media/video/s5p-fimc/fimc-capture.c @@ -350,7 +350,8 @@ static int queue_setup(struct vb2_queue *vq, const struct v4l2_format *pfmt, if (pixm) sizes[i] = max(size, pixm->plane_fmt[i].sizeimage); else - sizes[i] = size; + sizes[i] = max_t(u32, size, frame->payload[i]); + allocators[i] = ctx->fimc_dev->alloc_ctx; } @@ -924,10 +925,10 @@ static int fimc_capture_set_format(struct fimc_dev *fimc, struct v4l2_format *f) pix->width = mf->width; pix->height = mf->height; } + fimc_adjust_mplane_format(ff->fmt, pix->width, pix->height, pix); for (i = 0; i < ff->fmt->colplanes; i++) - ff->payload[i] = - (pix->width * pix->height * ff->fmt->depth[i]) / 8; + ff->payload[i] = pix->plane_fmt[i].sizeimage; set_frame_bounds(ff, pix->width, pix->height); /* Reset the composition rectangle if not yet configured */ diff --git a/drivers/media/video/s5p-fimc/fimc-core.c b/drivers/media/video/s5p-fimc/fimc-core.c index 92fc5a2..65124a2 100644 --- a/drivers/media/video/s5p-fimc/fimc-core.c +++ b/drivers/media/video/s5p-fimc/fimc-core.c @@ -741,8 +741,8 @@ void fimc_adjust_mplane_format(struct fimc_fmt *fmt, u32 width, u32 height, pix->width = width; for (i = 0; i < pix->num_planes; ++i) { - u32 bpl = pix->plane_fmt[i].bytesperline; - u32 *sizeimage = &pix->plane_fmt[i].sizeimage; + struct v4l2_plane_pix_format *plane_fmt = &pix->plane_fmt[i]; + u32 bpl = plane_fmt->bytesperline; if (fmt->colplanes > 1 && (bpl == 0 || bpl < pix->width)) bpl = pix->width; /* Planar */ @@ -754,8 +754,9 @@ void fimc_adjust_mplane_format(struct fimc_fmt *fmt, u32 width, u32 height, if (i == 0) /* Same bytesperline for each plane. */ bytesperline = bpl; - pix->plane_fmt[i].bytesperline = bytesperline; - *sizeimage = (pix->width * pix->height * fmt->depth[i]) / 8; + plane_fmt->bytesperline = bytesperline; + plane_fmt->sizeimage = max((pix->width * pix->height * + fmt->depth[i]) / 8, plane_fmt->sizeimage); } } -- cgit v0.10.2 From 0b4b1f199d367762ddb68cb3303431fa6c84a7d6 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Wed, 16 May 2012 15:03:50 -0300 Subject: [media] s5p-fimc: Remove superfluous checks for buffer type The checks are already done at the v4l2 framework. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-capture.c b/drivers/media/video/s5p-fimc/fimc-capture.c index 13df723..0fd12df 100644 --- a/drivers/media/video/s5p-fimc/fimc-capture.c +++ b/drivers/media/video/s5p-fimc/fimc-capture.c @@ -819,9 +819,6 @@ static int fimc_cap_g_fmt_mplane(struct file *file, void *fh, struct fimc_dev *fimc = video_drvdata(file); struct fimc_ctx *ctx = fimc->vid_cap.ctx; - if (f->type != V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE) - return -EINVAL; - return fimc_fill_format(&ctx->d_frame, f); } @@ -834,9 +831,6 @@ static int fimc_cap_try_fmt_mplane(struct file *file, void *fh, struct v4l2_mbus_framefmt mf; struct fimc_fmt *ffmt = NULL; - if (f->type != V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE) - return -EINVAL; - if (pix->pixelformat == V4L2_PIX_FMT_JPEG) { fimc_capture_try_format(ctx, &pix->width, &pix->height, NULL, &pix->pixelformat, @@ -888,8 +882,6 @@ static int fimc_capture_set_format(struct fimc_dev *fimc, struct v4l2_format *f) struct fimc_fmt *s_fmt = NULL; int ret, i; - if (f->type != V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE) - return -EINVAL; if (vb2_is_busy(&fimc->vid_cap.vbq)) return -EBUSY; -- cgit v0.10.2 From e3fc82e8b9f550d28f05600b98bcc8c26beef2ef Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Thu, 17 May 2012 14:22:10 -0300 Subject: [media] s5p-fimc: Prevent lock-up in multiple sensor systems The camera clocks managed by the driver were improperly reference counted and remained disabled when multiple video nodes were opened simultaneously. It manifested itself with following warning: [12.920000] WARNING: at drivers/media/video/s5p-fimc/fimc-mdevice.c:787 __fimc_md_set_camclk+0x1c0/0x1dc() [13.005000] Modules linked in: [13.005000] Backtrace: [13.040000] [] (dump_backtrace+0x0/0x10c) from [] (dump_stack+0x18/0x1c) [13.070000] r7:00000009 r6:00000313 r5:c02d576c r4:00000000 [13.155000] [] (dump_stack+0x0/0x1c) from [] (warn_slowpath_common+0x54/0x6c) [13.285000] [] (warn_slowpath_common+0x0/0x6c) from [] (warn_slowpath_null+0x24/0x2c) [13.360000] r9:e1981010 r8:00000000 r7:c061d3fc r6:e1981010 r5:e1981030 [13.430000] r4:00000000 [13.430000] [] (warn_slowpath_null+0x0/0x2c) from [] (__fimc_md_set_camclk+0x1c0/0x1dc) [13.550000] [] (__fimc_md_set_camclk+0x0/0x1dc) from [] (fimc_md_set_camclk+0x28/0x2c) [13.630000] [] (fimc_md_set_camclk+0x0/0x2c) from [] (__fimc_pipeline_shutdown+0x34/0x50) [13.705000] [] (__fimc_pipeline_shutdown+0x0/0x50) from [] (fimc_pipeline_shutdown+0x40/0x58) [13.765000] r5:e2391200 r4:e2357704 [13.805000] [] (fimc_pipeline_shutdown+0x0/0x58) from [] (fimc_capture_close+0xcc/0xe4) [13.915000] r5:e1b396c0 r4:e2357410 [13.915000] [] (fimc_capture_close+0x0/0xe4) from [] (v4l2_release+0x5c/0x80) [13.970000] r7:00000010 r6:e1d2d990 r5:e1b396c0 r4:e2394800 [14.000000] [] (v4l2_release+0x0/0x80) from [] (fput+0xc0/0x22c) [14.015000] r5:c157ef30 r4:e1b396c0 [14.015000] [] (fput+0x0/0x22c) from [] (filp_close+0x60/0x80) [14.080000] [] (filp_close+0x0/0x80) from [] (sys_close+0xb8/0xf4) [14.125000] r7:00000001 r6:e1b396c0 r5:c1400340 r4:c1400300 [14.125000] [] (sys_close+0x0/0xf4) from [] (ret_fast_syscall+0x0/0x30) [14.205000] r7:00000006 r6:beee5b94 r5:00000003 r4:b6f64fac Fix this, as well as potential memory leaks due to not calling v4l2_fh_release() on some error paths. Also remove some error logs printed for events that aren't critical and are normal conditions for some system configurations. Also check if the device have been properly run-time enabled during video node open. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-capture.c b/drivers/media/video/s5p-fimc/fimc-capture.c index 0fd12df..71e4838 100644 --- a/drivers/media/video/s5p-fimc/fimc-capture.c +++ b/drivers/media/video/s5p-fimc/fimc-capture.c @@ -480,37 +480,39 @@ static int fimc_capture_set_default_format(struct fimc_dev *fimc); static int fimc_capture_open(struct file *file) { struct fimc_dev *fimc = video_drvdata(file); - int ret = v4l2_fh_open(file); - - if (ret) - return ret; + int ret; dbg("pid: %d, state: 0x%lx", task_pid_nr(current), fimc->state); - /* Return if the corresponding video mem2mem node is already opened. */ if (fimc_m2m_active(fimc)) return -EBUSY; set_bit(ST_CAPT_BUSY, &fimc->state); - pm_runtime_get_sync(&fimc->pdev->dev); - - if (++fimc->vid_cap.refcnt == 1) { - ret = fimc_pipeline_initialize(&fimc->pipeline, - &fimc->vid_cap.vfd->entity, true); - if (ret < 0) { - dev_err(&fimc->pdev->dev, - "Video pipeline initialization failed\n"); - clear_bit(ST_CAPT_BUSY, &fimc->state); - pm_runtime_put_sync(&fimc->pdev->dev); - fimc->vid_cap.refcnt--; - v4l2_fh_release(file); - return ret; - } - ret = fimc_capture_ctrls_create(fimc); + ret = pm_runtime_get_sync(&fimc->pdev->dev); + if (ret < 0) + return ret; + + ret = v4l2_fh_open(file); + if (ret) + return ret; + + if (++fimc->vid_cap.refcnt != 1) + return 0; - if (!ret && !fimc->vid_cap.user_subdev_api) - ret = fimc_capture_set_default_format(fimc); + ret = fimc_pipeline_initialize(&fimc->pipeline, + &fimc->vid_cap.vfd->entity, true); + if (ret < 0) { + clear_bit(ST_CAPT_BUSY, &fimc->state); + pm_runtime_put_sync(&fimc->pdev->dev); + fimc->vid_cap.refcnt--; + v4l2_fh_release(file); + return ret; } + ret = fimc_capture_ctrls_create(fimc); + + if (!ret && !fimc->vid_cap.user_subdev_api) + ret = fimc_capture_set_default_format(fimc); + return ret; } diff --git a/drivers/media/video/s5p-fimc/fimc-lite.c b/drivers/media/video/s5p-fimc/fimc-lite.c index 400d701a..bbe93e4 100644 --- a/drivers/media/video/s5p-fimc/fimc-lite.c +++ b/drivers/media/video/s5p-fimc/fimc-lite.c @@ -451,21 +451,23 @@ static void fimc_lite_clear_event_counters(struct fimc_lite *fimc) static int fimc_lite_open(struct file *file) { struct fimc_lite *fimc = video_drvdata(file); - int ret = v4l2_fh_open(file); - - if (ret) - return ret; + int ret; set_bit(ST_FLITE_IN_USE, &fimc->state); - pm_runtime_get_sync(&fimc->pdev->dev); + ret = pm_runtime_get_sync(&fimc->pdev->dev); + if (ret < 0) + return ret; if (++fimc->ref_count != 1 || fimc->out_path != FIMC_IO_DMA) + return 0; + + ret = v4l2_fh_open(file); + if (ret < 0) return ret; ret = fimc_pipeline_initialize(&fimc->pipeline, &fimc->vfd->entity, true); if (ret < 0) { - v4l2_err(fimc->vfd, "Video pipeline initialization failed\n"); pm_runtime_put_sync(&fimc->pdev->dev); fimc->ref_count--; v4l2_fh_release(file); diff --git a/drivers/media/video/s5p-fimc/fimc-mdevice.c b/drivers/media/video/s5p-fimc/fimc-mdevice.c index dffe4da..52cef48 100644 --- a/drivers/media/video/s5p-fimc/fimc-mdevice.c +++ b/drivers/media/video/s5p-fimc/fimc-mdevice.c @@ -741,8 +741,8 @@ static void fimc_md_put_clocks(struct fimc_md *fmd) } static int __fimc_md_set_camclk(struct fimc_md *fmd, - struct fimc_sensor_info *s_info, - bool on) + struct fimc_sensor_info *s_info, + bool on) { struct s5p_fimc_isp_info *pdata = s_info->pdata; struct fimc_camclk_info *camclk; @@ -751,12 +751,10 @@ static int __fimc_md_set_camclk(struct fimc_md *fmd, if (WARN_ON(pdata->clk_id >= FIMC_MAX_CAMCLKS) || fmd == NULL) return -EINVAL; - if (s_info->clk_on == on) - return 0; camclk = &fmd->camclk[pdata->clk_id]; - dbg("camclk %d, f: %lu, clk: %p, on: %d", - pdata->clk_id, pdata->clk_frequency, camclk, on); + dbg("camclk %d, f: %lu, use_count: %d, on: %d", + pdata->clk_id, pdata->clk_frequency, camclk->use_count, on); if (on) { if (camclk->use_count > 0 && @@ -767,11 +765,9 @@ static int __fimc_md_set_camclk(struct fimc_md *fmd, clk_set_rate(camclk->clock, pdata->clk_frequency); camclk->frequency = pdata->clk_frequency; ret = clk_enable(camclk->clock); + dbg("Enabled camclk %d: f: %lu", pdata->clk_id, + clk_get_rate(camclk->clock)); } - s_info->clk_on = 1; - dbg("Enabled camclk %d: f: %lu", pdata->clk_id, - clk_get_rate(camclk->clock)); - return ret; } @@ -780,7 +776,6 @@ static int __fimc_md_set_camclk(struct fimc_md *fmd, if (--camclk->use_count == 0) { clk_disable(camclk->clock); - s_info->clk_on = 0; dbg("Disabled camclk %d", pdata->clk_id); } return ret; @@ -796,8 +791,6 @@ static int __fimc_md_set_camclk(struct fimc_md *fmd, * devices to which sensors can be attached, either directly or through * the MIPI CSI receiver. The clock is allowed here to be used by * multiple sensors concurrently if they use same frequency. - * The per sensor subdev clk_on attribute helps to synchronize accesses - * to the sclk_cam clocks from the video and media device nodes. * This function should only be called when the graph mutex is held. */ int fimc_md_set_camclk(struct v4l2_subdev *sd, bool on) diff --git a/drivers/media/video/s5p-fimc/fimc-mdevice.h b/drivers/media/video/s5p-fimc/fimc-mdevice.h index 3b8a349..1f5dbaf 100644 --- a/drivers/media/video/s5p-fimc/fimc-mdevice.h +++ b/drivers/media/video/s5p-fimc/fimc-mdevice.h @@ -47,7 +47,6 @@ struct fimc_camclk_info { * @pdata: sensor's atrributes passed as media device's platform data * @subdev: image sensor v4l2 subdev * @host: fimc device the sensor is currently linked to - * @clk_on: sclk_cam clock's state associated with this subdev * * This data structure applies to image sensor and the writeback subdevs. */ @@ -55,7 +54,6 @@ struct fimc_sensor_info { struct s5p_fimc_isp_info *pdata; struct v4l2_subdev *subdev; struct fimc_dev *host; - bool clk_on; }; /** -- cgit v0.10.2 From 316efab3e949e4fbdacb0975bf33adbad89115db Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Fri, 18 May 2012 13:31:28 -0300 Subject: [media] s5p-fimc: Fix fimc-lite system wide suspend procedure Only suspend the video pipeline devices if they were active before the pm.suspend() helper is called. This patch prevents following error: /# echo mem > /sys/power/state [ 34.965000] PM: Syncing filesystems ... done. [ 35.035000] Freezing user space processes ... (elapsed 0.01 seconds) done. ... [ 35.105000] dpm_run_callback(): platform_pm_suspend+0x0/0x5c returns -22 [ 35.105000] PM: Device exynos-fimc-lite.1 failed to suspend: error -22 [ 35.105000] PM: Some devices failed to suspend Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-lite.c b/drivers/media/video/s5p-fimc/fimc-lite.c index bbe93e4..b0a8ce6 100644 --- a/drivers/media/video/s5p-fimc/fimc-lite.c +++ b/drivers/media/video/s5p-fimc/fimc-lite.c @@ -1510,7 +1510,7 @@ static int fimc_lite_suspend(struct device *dev) return 0; ret = fimc_lite_stop_capture(fimc, suspend); - if (ret) + if (ret < 0 || !fimc_lite_active(fimc)) return ret; return fimc_pipeline_shutdown(&fimc->pipeline); -- cgit v0.10.2 From 0a198bcd511fcc60dc5daa203d221aafc95e0471 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Tue, 22 May 2012 13:50:14 -0300 Subject: [media] s5p-fimc: Shorten pixel formats description Shorten pixel format descriptions that exceed 32 characters so they're not being truncated when queried from user space. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-core.c b/drivers/media/video/s5p-fimc/fimc-core.c index 65124a2..987bff2 100644 --- a/drivers/media/video/s5p-fimc/fimc-core.c +++ b/drivers/media/video/s5p-fimc/fimc-core.c @@ -153,7 +153,7 @@ static struct fimc_fmt fimc_formats[] = { .colplanes = 2, .flags = FMT_FLAGS_M2M, }, { - .name = "YUV 4:2:0 non-contiguous 2-planar, Y/CbCr", + .name = "YUV 4:2:0 non-contig. 2p, Y/CbCr", .fourcc = V4L2_PIX_FMT_NV12M, .color = FIMC_FMT_YCBCR420, .depth = { 8, 4 }, @@ -161,7 +161,7 @@ static struct fimc_fmt fimc_formats[] = { .colplanes = 2, .flags = FMT_FLAGS_M2M, }, { - .name = "YUV 4:2:0 non-contiguous 3-planar, Y/Cb/Cr", + .name = "YUV 4:2:0 non-contig. 3p, Y/Cb/Cr", .fourcc = V4L2_PIX_FMT_YUV420M, .color = FIMC_FMT_YCBCR420, .depth = { 8, 2, 2 }, @@ -169,7 +169,7 @@ static struct fimc_fmt fimc_formats[] = { .colplanes = 3, .flags = FMT_FLAGS_M2M, }, { - .name = "YUV 4:2:0 non-contiguous 2-planar, Y/CbCr, tiled", + .name = "YUV 4:2:0 non-contig. 2p, tiled", .fourcc = V4L2_PIX_FMT_NV12MT, .color = FIMC_FMT_YCBCR420, .depth = { 8, 4 }, -- cgit v0.10.2 From 8183e7a7e2a4b542bd3044c5e3b0e116b0a7d2a1 Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Fri, 25 May 2012 07:04:01 -0300 Subject: [media] s5p-fimc: Update to the control handler lock changes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 77e7c4e624404c6edb5686b3d5f873c6008ed6b0 "v4l: Allow changing control handler lock" changed the lock field of struct v4l2_ctrl_handler to a pointer and this driver wasn't updated properly. This patch fixes following warning: drivers/media/video/s5p-fimc/fimc-core.c: In function ‘fimc_ctrls_activate’: drivers/media/video/s5p-fimc/fimc-core.c:644: warning: passing argument 1 of ‘mutex_lock’ from incompatible pointer type include/linux/mutex.h:152: note: expected ‘struct mutex *’ but argument is of type ‘struct mutex **’ drivers/media/video/s5p-fimc/fimc-core.c:663: warning: passing argument 1 of ‘mutex_unlock’ from incompatible pointer type include/linux/mutex.h:169: note: expected ‘struct mutex *’ but argument is of type ‘struct mutex **’ Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-core.c b/drivers/media/video/s5p-fimc/fimc-core.c index 987bff2..a4646ca 100644 --- a/drivers/media/video/s5p-fimc/fimc-core.c +++ b/drivers/media/video/s5p-fimc/fimc-core.c @@ -641,7 +641,7 @@ void fimc_ctrls_activate(struct fimc_ctx *ctx, bool active) if (!ctrls->ready) return; - mutex_lock(&ctrls->handler.lock); + mutex_lock(ctrls->handler.lock); v4l2_ctrl_activate(ctrls->rotate, active); v4l2_ctrl_activate(ctrls->hflip, active); v4l2_ctrl_activate(ctrls->vflip, active); @@ -660,7 +660,7 @@ void fimc_ctrls_activate(struct fimc_ctx *ctx, bool active) ctx->hflip = 0; ctx->vflip = 0; } - mutex_unlock(&ctrls->handler.lock); + mutex_unlock(ctrls->handler.lock); } /* Update maximum value of the alpha color control */ -- cgit v0.10.2 From a60a295986edcbca6603cd42b4e38d7c83d87fb6 Mon Sep 17 00:00:00 2001 From: Sakari Ailus Date: Fri, 25 May 2012 07:08:05 -0300 Subject: [media] s5p-fimc: media_entity_pipeline_start() may fail Take into account media_entity_pipeline_start() may fail. [s.nawrocki: rebased onto latest tree] Signed-off-by: Sakari Ailus Signed-off-by: Sylwester Nawrocki Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-capture.c b/drivers/media/video/s5p-fimc/fimc-capture.c index 71e4838..7ace324 100644 --- a/drivers/media/video/s5p-fimc/fimc-capture.c +++ b/drivers/media/video/s5p-fimc/fimc-capture.c @@ -1045,8 +1045,10 @@ static int fimc_cap_streamon(struct file *file, void *priv, if (fimc_capture_active(fimc)) return -EBUSY; - media_entity_pipeline_start(&p->subdevs[IDX_SENSOR]->entity, + ret = media_entity_pipeline_start(&p->subdevs[IDX_SENSOR]->entity, p->m_pipeline); + if (ret < 0) + return ret; if (fimc->vid_cap.user_subdev_api) { ret = fimc_pipeline_validate(fimc); -- cgit v0.10.2 From a1a5861bd9ca3a6cb7ab89dd836da8cd158986b0 Mon Sep 17 00:00:00 2001 From: Sachin Kamat Date: Fri, 25 May 2012 03:29:38 -0300 Subject: [media] s5p-fimc: Fix compiler warning in fimc-lite.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch is an update to changed media_entity_pipeline_start() signature in commit af88be3887c1a0b20d0792c3c237a67c73ef3286, "media: Add link_validate() op to check links to the sink pad" It fixes the following warning: drivers/media/video/s5p-fimc/fimc-lite.c: In function ‘fimc_lite_streamon’: drivers/media/video/s5p-fimc/fimc-lite.c:765:29: warning: ignoring return value of ‘media_entity_pipeline_start’, declared with attribute warn_unused_result [-Wunused-result] Signed-off-by: Sachin Kamat Signed-off-by: Sylwester Nawrocki Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-lite.c b/drivers/media/video/s5p-fimc/fimc-lite.c index b0a8ce6..4d269b8 100644 --- a/drivers/media/video/s5p-fimc/fimc-lite.c +++ b/drivers/media/video/s5p-fimc/fimc-lite.c @@ -764,7 +764,9 @@ static int fimc_lite_streamon(struct file *file, void *priv, if (fimc_lite_active(fimc)) return -EBUSY; - media_entity_pipeline_start(&sensor->entity, p->m_pipeline); + ret = media_entity_pipeline_start(&sensor->entity, p->m_pipeline); + if (ret < 0) + return ret; ret = fimc_pipeline_validate(fimc); if (ret) { -- cgit v0.10.2 From f676fa068819357f716953920b764554fe116b3c Mon Sep 17 00:00:00 2001 From: Sachin Kamat Date: Fri, 25 May 2012 03:29:40 -0300 Subject: [media] s5p-fimc: Stop media entity pipeline if fimc_pipeline_validate fails Stops the media entity pipeline which was started earlier if fimc_pipeline_validate fails. [s.nawrocki: reworked to not exceed 80 characters line length] Signed-off-by: Sachin Kamat Signed-off-by: Sylwester Nawrocki Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-capture.c b/drivers/media/video/s5p-fimc/fimc-capture.c index 7ace324..725812a 100644 --- a/drivers/media/video/s5p-fimc/fimc-capture.c +++ b/drivers/media/video/s5p-fimc/fimc-capture.c @@ -1040,20 +1040,22 @@ static int fimc_cap_streamon(struct file *file, void *priv, { struct fimc_dev *fimc = video_drvdata(file); struct fimc_pipeline *p = &fimc->pipeline; + struct v4l2_subdev *sd = p->subdevs[IDX_SENSOR]; int ret; if (fimc_capture_active(fimc)) return -EBUSY; - ret = media_entity_pipeline_start(&p->subdevs[IDX_SENSOR]->entity, - p->m_pipeline); + ret = media_entity_pipeline_start(&sd->entity, p->m_pipeline); if (ret < 0) return ret; if (fimc->vid_cap.user_subdev_api) { ret = fimc_pipeline_validate(fimc); - if (ret) + if (ret < 0) { + media_entity_pipeline_stop(&sd->entity); return ret; + } } return vb2_streamon(&fimc->vid_cap.vbq, type); } -- cgit v0.10.2 From 4069d6f86bebce1a1e3456ef721511b4b81958f8 Mon Sep 17 00:00:00 2001 From: Mattia Dongili Date: Sat, 9 Jun 2012 13:18:08 +0900 Subject: sony-laptop: use an enum for SNC event types Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index 210d4ae..615ab06 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -1172,6 +1172,10 @@ static int sony_nc_hotkeys_decode(u32 event, unsigned int handle) /* * ACPI callbacks */ +enum event_types { + HOTKEY = 1, + KILLSWITCH +}; static void sony_nc_notify(struct acpi_device *device, u32 event) { u32 real_ev = event; @@ -1196,7 +1200,7 @@ static void sony_nc_notify(struct acpi_device *device, u32 event) /* hotkey event */ case 0x0100: case 0x0127: - ev_type = 1; + ev_type = HOTKEY; real_ev = sony_nc_hotkeys_decode(event, handle); if (real_ev > 0) @@ -1216,7 +1220,7 @@ static void sony_nc_notify(struct acpi_device *device, u32 event) * update the rfkill device status when the * switch is moved. */ - ev_type = 2; + ev_type = KILLSWITCH; sony_call_snc_handle(handle, 0x0100, &result); real_ev = result & 0x03; @@ -1238,7 +1242,7 @@ static void sony_nc_notify(struct acpi_device *device, u32 event) } else { /* old style event */ - ev_type = 1; + ev_type = HOTKEY; sony_laptop_report_input_event(real_ev); } -- cgit v0.10.2 From bb384b5295323ed58260aeaff22d8bbe32988396 Mon Sep 17 00:00:00 2001 From: Marco Chiappero Date: Sat, 9 Jun 2012 13:18:09 +0900 Subject: sony-laptop: notify userspace of GFX switch position changes Some Vaios come with both integrated and discrete graphics, plus a switch for choosing one of the two. When the switch position is changed, a notification is generated. Signed-off-by: Marco Chiappero Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index 615ab06..4f42e56 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -1174,7 +1174,8 @@ static int sony_nc_hotkeys_decode(u32 event, unsigned int handle) */ enum event_types { HOTKEY = 1, - KILLSWITCH + KILLSWITCH, + GFX_SWITCH }; static void sony_nc_notify(struct acpi_device *device, u32 event) { @@ -1230,6 +1231,24 @@ static void sony_nc_notify(struct acpi_device *device, u32 event) break; + case 0x0128: + case 0x0146: + /* Hybrid GFX switching */ + sony_call_snc_handle(handle, 0x0000, &result); + dprintk("GFX switch event received (reason: %s)\n", + (result & 0x01) ? + "switch change" : "unknown"); + + /* verify the switch state + * 1: discrete GFX + * 0: integrated GFX + */ + sony_call_snc_handle(handle, 0x0100, &result); + + ev_type = GFX_SWITCH; + real_ev = result & 0xff; + break; + default: dprintk("Unknown event 0x%x for handle 0x%x\n", event, handle); -- cgit v0.10.2 From 15aa5c75468a103cdee1a0e0ec26aad979bf71a5 Mon Sep 17 00:00:00 2001 From: Mattia Dongili Date: Mon, 11 Jun 2012 07:18:31 +0900 Subject: sony-laptop: store battery care limits on batteries Some models offer the option to store the limits on the battery (firmware?). Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index 4f42e56..f045e3e 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -1916,32 +1916,33 @@ static ssize_t sony_nc_battery_care_limit_store(struct device *dev, * bits 4,5: store the limit into the EC * bits 6,7: store the limit into the battery */ + cmd = 0; - /* - * handle 0x0115 should allow storing on battery too; - * handle 0x0136 same as 0x0115 + health status; - * handle 0x013f, same as 0x0136 but no storing on the battery - * - * Store only inside the EC for now, regardless the handle number - */ - if (value == 0) - /* disable limits */ - cmd = 0x0; + if (value > 0) { + if (value <= 50) + cmd = 0x20; - else if (value <= 50) - cmd = 0x21; + else if (value <= 80) + cmd = 0x10; - else if (value <= 80) - cmd = 0x11; + else if (value <= 100) + cmd = 0x30; + + else + return -EINVAL; - else if (value <= 100) - cmd = 0x31; + /* + * handle 0x0115 should allow storing on battery too; + * handle 0x0136 same as 0x0115 + health status; + * handle 0x013f, same as 0x0136 but no storing on the battery + */ + if (bcare_ctl->handle != 0x013f) + cmd = cmd | (cmd << 2); - else - return -EINVAL; + cmd = (cmd | 0x1) << 0x10; + } - if (sony_call_snc_handle(bcare_ctl->handle, (cmd << 0x10) | 0x0100, - &result)) + if (sony_call_snc_handle(bcare_ctl->handle, cmd | 0x0100, &result)) return -EIO; return count; -- cgit v0.10.2 From 014fc8fbece33d42e2aa92d289fffa213a159321 Mon Sep 17 00:00:00 2001 From: Mattia Dongili Date: Sat, 9 Jun 2012 13:18:11 +0900 Subject: sony-laptop: add lid backlight support for handle 0x143 Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index f045e3e..a94e13f 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -1010,6 +1010,7 @@ static ssize_t sony_nc_sysfs_store(struct device *dev, struct sony_backlight_props { struct backlight_device *dev; int handle; + int cmd_base; u8 offset; u8 maxlvl; }; @@ -1037,7 +1038,7 @@ static int sony_nc_get_brightness_ng(struct backlight_device *bd) struct sony_backlight_props *sdev = (struct sony_backlight_props *)bl_get_data(bd); - sony_call_snc_handle(sdev->handle, 0x0200, &result); + sony_call_snc_handle(sdev->handle, sdev->cmd_base + 0x100, &result); return (result & 0xff) - sdev->offset; } @@ -1049,7 +1050,8 @@ static int sony_nc_update_status_ng(struct backlight_device *bd) (struct sony_backlight_props *)bl_get_data(bd); value = bd->props.brightness + sdev->offset; - if (sony_call_snc_handle(sdev->handle, 0x0100 | (value << 16), &result)) + if (sony_call_snc_handle(sdev->handle, sdev->cmd_base | (value << 0x10), + &result)) return -EIO; return value; @@ -2496,6 +2498,7 @@ static void sony_nc_backlight_ng_read_limits(int handle, { u64 offset; int i; + int lvl_table_len = 0; u8 min = 0xff, max = 0x00; unsigned char buffer[32] = { 0 }; @@ -2515,11 +2518,21 @@ static void sony_nc_backlight_ng_read_limits(int handle, if (i < 0) return; + switch (handle) { + case 0x012f: + case 0x0137: + lvl_table_len = 9; + break; + case 0x143: + lvl_table_len = 16; + break; + } + /* the buffer lists brightness levels available, brightness levels are * from position 0 to 8 in the array, other values are used by ALS * control. */ - for (i = 0; i < 9 && i < ARRAY_SIZE(buffer); i++) { + for (i = 0; i < lvl_table_len && i < ARRAY_SIZE(buffer); i++) { dprintk("Brightness level: %d\n", buffer[i]); @@ -2546,14 +2559,22 @@ static void sony_nc_backlight_setup(void) if (sony_find_snc_handle(0x12f) != -1) { ops = &sony_backlight_ng_ops; + sony_bl_props.cmd_base = 0x0100; sony_nc_backlight_ng_read_limits(0x12f, &sony_bl_props); max_brightness = sony_bl_props.maxlvl - sony_bl_props.offset; } else if (sony_find_snc_handle(0x137) != -1) { ops = &sony_backlight_ng_ops; + sony_bl_props.cmd_base = 0x0100; sony_nc_backlight_ng_read_limits(0x137, &sony_bl_props); max_brightness = sony_bl_props.maxlvl - sony_bl_props.offset; + } else if (sony_find_snc_handle(0x143) != -1) { + ops = &sony_backlight_ng_ops; + sony_bl_props.cmd_base = 0x3000; + sony_nc_backlight_ng_read_limits(0x143, &sony_bl_props); + max_brightness = sony_bl_props.maxlvl - sony_bl_props.offset; + } else if (ACPI_SUCCESS(acpi_get_handle(sony_nc_acpi_handle, "GBRT", &unused))) { ops = &sony_backlight_ops; -- cgit v0.10.2 From ca3c2c706de39b3400e57254dce054bf7350efa2 Mon Sep 17 00:00:00 2001 From: Mattia Dongili Date: Sat, 9 Jun 2012 13:18:12 +0900 Subject: sony-laptop: input initialization should be done before SNC SNC needs input devices so better have those ready before starting handle events. Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index a94e13f..89ff6d8 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -2642,6 +2642,12 @@ static int sony_nc_add(struct acpi_device *device) } } + result = sony_laptop_setup_input(device); + if (result) { + pr_err("Unable to create input devices\n"); + goto outplatform; + } + if (ACPI_SUCCESS(acpi_get_handle(sony_nc_acpi_handle, "ECON", &handle))) { int arg = 1; @@ -2659,12 +2665,6 @@ static int sony_nc_add(struct acpi_device *device) } /* setup input devices and helper fifo */ - result = sony_laptop_setup_input(device); - if (result) { - pr_err("Unable to create input devices\n"); - goto outsnc; - } - if (acpi_video_backlight_support()) { pr_info("brightness ignored, must be controlled by ACPI video driver\n"); } else { @@ -2712,22 +2712,21 @@ static int sony_nc_add(struct acpi_device *device) return 0; - out_sysfs: +out_sysfs: for (item = sony_nc_values; item->name; ++item) { device_remove_file(&sony_pf_device->dev, &item->devattr); } sony_nc_backlight_cleanup(); - - sony_laptop_remove_input(); - - outsnc: sony_nc_function_cleanup(sony_pf_device); sony_nc_handles_cleanup(sony_pf_device); - outpresent: +outplatform: + sony_laptop_remove_input(); + +outpresent: sony_pf_remove(); - outwalk: +outwalk: sony_nc_rfkill_cleanup(); return result; } -- cgit v0.10.2 From c7a2918373983b32db3ca35823d930641747e26f Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 9 Jun 2012 13:18:13 +0900 Subject: sony-laptop: fix sony_nc_sysfs_store() We made this an unsigned long and it causes a bug on 64 bit big endian systems when we try to pass the value to sony_nc_int_call(). Also value has to be signed because validate() returns negative error codes. Signed-off-by: Dan Carpenter Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index 89ff6d8..78e6389 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -973,7 +973,7 @@ static ssize_t sony_nc_sysfs_store(struct device *dev, struct device_attribute *attr, const char *buffer, size_t count) { - unsigned long value = 0; + int value; int ret = 0; struct sony_nc_value *item = container_of(attr, struct sony_nc_value, devattr); @@ -984,7 +984,7 @@ static ssize_t sony_nc_sysfs_store(struct device *dev, if (count > 31) return -EINVAL; - if (kstrtoul(buffer, 10, &value)) + if (kstrtoint(buffer, 10, &value)) return -EINVAL; if (item->validate) @@ -994,7 +994,7 @@ static ssize_t sony_nc_sysfs_store(struct device *dev, return value; ret = sony_nc_int_call(sony_nc_acpi_handle, *item->acpiset, - (int *)&value, NULL); + &value, NULL); if (ret < 0) return -EIO; -- cgit v0.10.2 From 56f4a9f76d8ce7c7cef92905c909aad0c7e5c9db Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 9 Jun 2012 13:18:14 +0900 Subject: sony-laptop: fix a couple signedness bugs This needs to be signed to handle negative error codes. Remove a redundant check, read_limits is always called with a valid handle. Signed-off-by: Dan Carpenter Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index 78e6389..b9499b6 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -2139,7 +2139,7 @@ static ssize_t sony_nc_thermal_mode_show(struct device *dev, struct device_attribute *attr, char *buffer) { ssize_t count = 0; - unsigned int mode = sony_nc_thermal_mode_get(); + int mode = sony_nc_thermal_mode_get(); if (mode < 0) return mode; @@ -2507,8 +2507,6 @@ static void sony_nc_backlight_ng_read_limits(int handle, props->maxlvl = 0xff; offset = sony_find_snc_handle(handle); - if (offset < 0) - return; /* try to read the boundaries from ACPI tables, if we fail the above * defaults should be reasonable -- cgit v0.10.2 From a1071a5abf1f4d4a7f8324e21b8b32b1342013c7 Mon Sep 17 00:00:00 2001 From: Mattia Dongili Date: Thu, 14 Jun 2012 06:36:02 +0900 Subject: sony-laptop: correct find_snc_handle failure checks Since bab7084c745bf4d75b760728387f375fd34dc683, find_snc_handle returns -EINVAL, not -1. Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index b9499b6..d456ff0 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -2555,19 +2555,19 @@ static void sony_nc_backlight_setup(void) const struct backlight_ops *ops = NULL; struct backlight_properties props; - if (sony_find_snc_handle(0x12f) != -1) { + if (sony_find_snc_handle(0x12f) >= 0) { ops = &sony_backlight_ng_ops; sony_bl_props.cmd_base = 0x0100; sony_nc_backlight_ng_read_limits(0x12f, &sony_bl_props); max_brightness = sony_bl_props.maxlvl - sony_bl_props.offset; - } else if (sony_find_snc_handle(0x137) != -1) { + } else if (sony_find_snc_handle(0x137) >= 0) { ops = &sony_backlight_ng_ops; sony_bl_props.cmd_base = 0x0100; sony_nc_backlight_ng_read_limits(0x137, &sony_bl_props); max_brightness = sony_bl_props.maxlvl - sony_bl_props.offset; - } else if (sony_find_snc_handle(0x143) != -1) { + } else if (sony_find_snc_handle(0x143) >= 0) { ops = &sony_backlight_ng_ops; sony_bl_props.cmd_base = 0x3000; sony_nc_backlight_ng_read_limits(0x143, &sony_bl_props); -- cgit v0.10.2 From 57f9616b79549e772cf4dd3aa1d2df5b6c8acdfa Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 12 Jun 2012 19:28:50 +0300 Subject: ideapad: uninitialized data in ideapad_acpi_add() We only initialize the high bits of "cfg". It probably doesn't cause a problem given that this is platform specific code and doesn't have to worry about endianness etc. But it's sort of messy. Signed-off-by: Dan Carpenter Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/ideapad-laptop.c b/drivers/platform/x86/ideapad-laptop.c index 4f20f8d..17f6dfd 100644 --- a/drivers/platform/x86/ideapad-laptop.c +++ b/drivers/platform/x86/ideapad-laptop.c @@ -694,10 +694,10 @@ MODULE_DEVICE_TABLE(acpi, ideapad_device_ids); static int __devinit ideapad_acpi_add(struct acpi_device *adevice) { int ret, i; - unsigned long cfg; + int cfg; struct ideapad_private *priv; - if (read_method_int(adevice->handle, "_CFG", (int *)&cfg)) + if (read_method_int(adevice->handle, "_CFG", &cfg)) return -ENODEV; priv = kzalloc(sizeof(*priv), GFP_KERNEL); @@ -721,7 +721,7 @@ static int __devinit ideapad_acpi_add(struct acpi_device *adevice) goto input_failed; for (i = 0; i < IDEAPAD_RFKILL_DEV_NUM; i++) { - if (test_bit(ideapad_rfk_data[i].cfgbit, &cfg)) + if (test_bit(ideapad_rfk_data[i].cfgbit, &priv->cfg)) ideapad_register_rfkill(adevice, i); else priv->rfk[i] = NULL; -- cgit v0.10.2 From 88ca518b0bb4161e5f20f8a1d9cc477cae294e54 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 25 Jun 2012 15:07:17 +0200 Subject: intel_ips: blacklist HP ProBook laptops intel_ips driver spews the warning message "ME failed to update for more than 1s, likely hung" at each second endlessly on HP ProBook laptops with IronLake. As this has never worked, better to blacklist the driver for now. Cc: Signed-off-by: Takashi Iwai Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/intel_ips.c b/drivers/platform/x86/intel_ips.c index 0ffdb3c..9af4257 100644 --- a/drivers/platform/x86/intel_ips.c +++ b/drivers/platform/x86/intel_ips.c @@ -72,6 +72,7 @@ #include #include #include +#include #include #include #include @@ -1485,6 +1486,24 @@ static DEFINE_PCI_DEVICE_TABLE(ips_id_table) = { MODULE_DEVICE_TABLE(pci, ips_id_table); +static int ips_blacklist_callback(const struct dmi_system_id *id) +{ + pr_info("Blacklisted intel_ips for %s\n", id->ident); + return 1; +} + +static const struct dmi_system_id ips_blacklist[] = { + { + .callback = ips_blacklist_callback, + .ident = "HP ProBook", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"), + DMI_MATCH(DMI_PRODUCT_NAME, "HP ProBook"), + }, + }, + { } /* terminating entry */ +}; + static int ips_probe(struct pci_dev *dev, const struct pci_device_id *id) { u64 platform_info; @@ -1494,6 +1513,9 @@ static int ips_probe(struct pci_dev *dev, const struct pci_device_id *id) u16 htshi, trc, trc_required_mask; u8 tse; + if (dmi_check_system(ips_blacklist)) + return -ENODEV; + ips = kzalloc(sizeof(struct ips_driver), GFP_KERNEL); if (!ips) return -ENOMEM; -- cgit v0.10.2 From 9e303f228c24d529bf479196d290eedc2a04cd5e Mon Sep 17 00:00:00 2001 From: Grazvydas Ignotas Date: Sat, 16 Jun 2012 22:01:25 +0300 Subject: gpio/omap: fix irq loss while in idle with debounce on It seems that currently GPIO module is not working correctly during idle when debounce is enabled - the system almost never responds to button presses (observed on OMAP3530 ES2.1 and OMAP3630 ES1.2 pandora boards). Even though wakeups are probably working, it seems that the GPIO module itself is unable to detect input events and generate interrupts. OMAP35x TRM also states that: "If the debounce clock is inactive, the debounce cell gates all input signals and thus cannot be used." So whenever we are disabling debounce clocks (for PM or other reasons), be sure the module's debounce feature is disabled too. Cc: Kevin Hilman Signed-off-by: Grazvydas Ignotas Signed-off-by: Kevin Hilman diff --git a/drivers/gpio/gpio-omap.c b/drivers/gpio/gpio-omap.c index c4ed172..ff213e7 100644 --- a/drivers/gpio/gpio-omap.c +++ b/drivers/gpio/gpio-omap.c @@ -174,12 +174,22 @@ static inline void _gpio_dbck_enable(struct gpio_bank *bank) if (bank->dbck_enable_mask && !bank->dbck_enabled) { clk_enable(bank->dbck); bank->dbck_enabled = true; + + __raw_writel(bank->dbck_enable_mask, + bank->base + bank->regs->debounce_en); } } static inline void _gpio_dbck_disable(struct gpio_bank *bank) { if (bank->dbck_enable_mask && bank->dbck_enabled) { + /* + * Disable debounce before cutting it's clock. If debounce is + * enabled but the clock is not, GPIO module seems to be unable + * to detect events and generate interrupts at least on OMAP3. + */ + __raw_writel(0, bank->base + bank->regs->debounce_en); + clk_disable(bank->dbck); bank->dbck_enabled = false; } -- cgit v0.10.2 From 8ca78f3eda4bf1799e8c4ba02035623fd7a347df Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Wed, 27 Jun 2012 15:05:48 +0200 Subject: Btrfs: avoid waiting for delayed refs when we must not We track two conditions to decide if we should sleep while waiting for more delayed refs, the number of delayed refs (num_refs) and the first entry in the list of blockers (first_seq). When we suspect staleness, we save num_refs and do one more cycle. If nothing changes, we then save first_seq for later comparison and do wait_event. We ought to save first_seq the very same moment we're saving num_refs. Otherwise we cannot be sure that nothing has changed and we might start waiting when we shouldn't, which could lead to starvation. Signed-off-by: Jan Schmidt diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 4b5a1e1..6e1d367 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2347,12 +2347,10 @@ next: return count; } - static void wait_for_more_refs(struct btrfs_delayed_ref_root *delayed_refs, - unsigned long num_refs) + unsigned long num_refs, + struct list_head *first_seq) { - struct list_head *first_seq = delayed_refs->seq_head.next; - spin_unlock(&delayed_refs->lock); pr_debug("waiting for more refs (num %ld, first %p)\n", num_refs, first_seq); @@ -2381,6 +2379,7 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_root *delayed_refs; struct btrfs_delayed_ref_node *ref; struct list_head cluster; + struct list_head *first_seq = NULL; int ret; u64 delayed_start; int run_all = count == (unsigned long)-1; @@ -2436,8 +2435,10 @@ again: */ consider_waiting = 1; num_refs = delayed_refs->num_entries; + first_seq = root->fs_info->tree_mod_seq_list.next; } else { - wait_for_more_refs(delayed_refs, num_refs); + wait_for_more_refs(delayed_refs, + num_refs, first_seq); /* * after waiting, things have changed. we * dropped the lock and someone else might have -- cgit v0.10.2 From 9345457f4a539a40056431aeb6f068750857472f Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Wed, 27 Jun 2012 15:23:09 +0200 Subject: Btrfs: support root level changes in __resolve_indirect_ref With the tree mod log, we can have a tree that's two levels high, but btrfs_search_old_slot may still return a path with the tree root at level one instead. __resolve_indirect_ref must care for this and accept parents in a lower level than expected. Signed-off-by: Jan Schmidt diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index 7301cdb..cf0df90 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -301,10 +301,14 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info, goto out; eb = path->nodes[level]; - if (!eb) { - WARN_ON(1); - ret = 1; - goto out; + while (!eb) { + if (!level) { + WARN_ON(1); + ret = 1; + goto out; + } + level--; + eb = path->nodes[level]; } ret = add_all_parents(root, path, parents, level, &ref->key_for_search, -- cgit v0.10.2 From 28da9fb4467f7a650cd31af6dfad3a4e4a3abf6e Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Thu, 21 Jun 2012 10:59:13 +0200 Subject: Btrfs: fix tree mod log for root replacements at leaf level For the tree mod log, we don't log any operations at leaf level. If the root is at the leaf level (i.e. the tree consists only of the root), then __tree_mod_log_oldest_root will find a ROOT_REPLACE operation in the log (because we always log that one no matter which level), but no other operations. With this patch __tree_mod_log_oldest_root exits cleanly instead of BUGging in this situation. get_old_root checks if its really a root at leaf level in case we don't have any operations and WARNs if this assumption breaks. Signed-off-by: Jan Schmidt diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 15cbc2b..7d1e4fc 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1024,11 +1024,18 @@ __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info, if (!looped && !tm) return 0; /* - * we must have key remove operations in the log before the - * replace operation. + * if there are no tree operation for the oldest root, we simply + * return it. this should only happen if that (old) root is at + * level 0. */ - BUG_ON(!tm); + if (!tm) + break; + /* + * if there's an operation that's not a root replacement, we + * found the oldest version of our root. normally, we'll find a + * MOD_LOG_KEY_REMOVE_WHILE_FREEING operation here. + */ if (tm->op != MOD_LOG_ROOT_REPLACE) break; @@ -1192,16 +1199,8 @@ get_old_root(struct btrfs_root *root, u64 time_seq) } tm = tree_mod_log_search(root->fs_info, logical, time_seq); - /* - * there was an item in the log when __tree_mod_log_oldest_root - * returned. this one must not go away, because the time_seq passed to - * us must be blocking its removal. - */ - BUG_ON(!tm); - if (old_root) - eb = alloc_dummy_extent_buffer(tm->index << PAGE_CACHE_SHIFT, - root->nodesize); + eb = alloc_dummy_extent_buffer(logical, root->nodesize); else eb = btrfs_clone_extent_buffer(root->node); btrfs_tree_read_unlock(root->node); @@ -1216,7 +1215,10 @@ get_old_root(struct btrfs_root *root, u64 time_seq) btrfs_set_header_level(eb, old_root->level); btrfs_set_header_generation(eb, old_generation); } - __tree_mod_log_rewind(eb, time_seq, tm); + if (tm) + __tree_mod_log_rewind(eb, time_seq, tm); + else + WARN_ON(btrfs_header_level(eb) != 0); extent_buffer_get(eb); return eb; -- cgit v0.10.2 From c3e0696523862c48b4d8c73ffb2867e9db478338 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Thu, 21 Jun 2012 11:01:06 +0200 Subject: Btrfs: always put insert_ptr modifications into the tree mod log Several callers of insert_ptr set the tree_mod_log parameter to 0 to avoid addition to the tree mod log. In fact, we need all of those operations. This commit simply removes the additional parameter and makes addition to the tree mod log unconditional. Signed-off-by: Jan Schmidt diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 7d1e4fc..e005d9b 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -2997,7 +2997,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans, static void insert_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, struct btrfs_disk_key *key, u64 bytenr, - int slot, int level, int tree_mod_log) + int slot, int level) { struct extent_buffer *lower; int nritems; @@ -3010,7 +3010,7 @@ static void insert_ptr(struct btrfs_trans_handle *trans, BUG_ON(slot > nritems); BUG_ON(nritems == BTRFS_NODEPTRS_PER_BLOCK(root)); if (slot != nritems) { - if (tree_mod_log && level) + if (level) tree_mod_log_eb_move(root->fs_info, lower, slot + 1, slot, nritems - slot); memmove_extent_buffer(lower, @@ -3018,7 +3018,7 @@ static void insert_ptr(struct btrfs_trans_handle *trans, btrfs_node_key_ptr_offset(slot), (nritems - slot) * sizeof(struct btrfs_key_ptr)); } - if (tree_mod_log && level) { + if (level) { ret = tree_mod_log_insert_key(root->fs_info, lower, slot, MOD_LOG_KEY_ADD); BUG_ON(ret < 0); @@ -3106,7 +3106,7 @@ static noinline int split_node(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(split); insert_ptr(trans, root, path, &disk_key, split->start, - path->slots[level + 1] + 1, level + 1, 1); + path->slots[level + 1] + 1, level + 1); if (path->slots[level] >= mid) { path->slots[level] -= mid; @@ -3643,7 +3643,7 @@ static noinline void copy_for_split(struct btrfs_trans_handle *trans, btrfs_set_header_nritems(l, mid); btrfs_item_key(right, &disk_key, 0); insert_ptr(trans, root, path, &disk_key, right->start, - path->slots[1] + 1, 1, 0); + path->slots[1] + 1, 1); btrfs_mark_buffer_dirty(right); btrfs_mark_buffer_dirty(l); @@ -3850,7 +3850,7 @@ again: if (mid <= slot) { btrfs_set_header_nritems(right, 0); insert_ptr(trans, root, path, &disk_key, right->start, - path->slots[1] + 1, 1, 0); + path->slots[1] + 1, 1); btrfs_tree_unlock(path->nodes[0]); free_extent_buffer(path->nodes[0]); path->nodes[0] = right; @@ -3859,7 +3859,7 @@ again: } else { btrfs_set_header_nritems(right, 0); insert_ptr(trans, root, path, &disk_key, right->start, - path->slots[1], 1, 0); + path->slots[1], 1); btrfs_tree_unlock(path->nodes[0]); free_extent_buffer(path->nodes[0]); path->nodes[0] = right; -- cgit v0.10.2 From 155725c9c051a343be5e555bf943da827e6cf721 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Fri, 22 Jun 2012 14:01:00 +0200 Subject: Btrfs: leave critical region in btrfs_find_all_roots as soon as possible When delayed refs exist, btrfs_find_all_roots used to hold the delayed ref mutex way longer than actually required. We ought to drop it immediately after we're done collecting all the delayed refs. Signed-off-by: Jan Schmidt diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index cf0df90..a383c18 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -839,6 +839,7 @@ again: } ret = __add_delayed_refs(head, delayed_ref_seq, &prefs_delayed); + mutex_unlock(&head->mutex); if (ret) { spin_unlock(&delayed_refs->lock); goto out; @@ -932,8 +933,6 @@ again: } out: - if (head) - mutex_unlock(&head->mutex); btrfs_free_path(path); while (!list_empty(&prefs)) { ref = list_first_entry(&prefs, struct __prelim_ref, list); -- cgit v0.10.2 From 19956c7e94a7a58d6df8c4db5ae62f9109a7c663 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Fri, 22 Jun 2012 14:52:13 +0200 Subject: Btrfs: fix tree mod log rewind of ADD operations When a MOD_LOG_KEY_ADD operation is rewinded, we remove the key from the tree block. If its not the last key, removal involves a move operation. This move operation was explicitly done before this commit. However, at insertion time, there's a move operation before the actual addition to make room for the new key, which is recorded in the tree mod log as well. This means, we must drop the move operation when rewinding the add operation, because the next operation we'll be rewinding will be the corresponding MOD_LOG_MOVE_KEYS operation. Signed-off-by: Jan Schmidt diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index e005d9b..b98f860 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1094,11 +1094,7 @@ __tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq, tm->generation); break; case MOD_LOG_KEY_ADD: - if (tm->slot != n - 1) { - o_dst = btrfs_node_key_ptr_offset(tm->slot); - o_src = btrfs_node_key_ptr_offset(tm->slot + 1); - memmove_extent_buffer(eb, o_dst, o_src, p_size); - } + /* if a move operation is needed it's in the log */ n--; break; case MOD_LOG_MOVE_KEYS: -- cgit v0.10.2 From d42244a0d36ad0939c5f173ebf15841a0678899c Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Fri, 22 Jun 2012 14:51:15 +0200 Subject: Btrfs: resolve tree mod log locking issue in btrfs_next_leaf With the tree mod log, we may end up with two roots (the current root and a rewinded version of it) both pointing to two leaves, l1 and l2, of which l2 had already been cow-ed in the current transaction. If we don't rewind any tree blocks, we cannot have two roots both pointing to an already cowed tree block. Now there is btrfs_next_leaf, which has a leaf locked and wants a lock on the next (right) leaf. And there is push_leaf_left, which has a (cowed!) leaf locked and wants a lock on the previous (left) leaf. In order to solve this dead lock situation, we use try_lock in btrfs_next_leaf (only in case it's called with a tree mod log time_seq paramter) and if we fail to get a lock on the next leaf, we give up our lock on the current leaf and retry from the very beginning. Signed-off-by: Jan Schmidt diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index b98f860..8206b39 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -5119,6 +5119,18 @@ again: if (!path->skip_locking) { ret = btrfs_try_tree_read_lock(next); + if (!ret && time_seq) { + /* + * If we don't get the lock, we may be racing + * with push_leaf_left, holding that lock while + * itself waiting for the leaf we've currently + * locked. To solve this situation, we give up + * on our lock and cycle. + */ + btrfs_release_path(path); + cond_resched(); + goto again; + } if (!ret) { btrfs_set_path_blocking(path); btrfs_tree_read_lock(next); -- cgit v0.10.2 From 61c91dd4a58b21a783e37208f4d02e3cb4b637c4 Mon Sep 17 00:00:00 2001 From: Ping Cheng Date: Thu, 28 Jun 2012 16:46:27 -0700 Subject: Input: wacom - fix retrieving touch_max bug rep_data is not an array anymore, so taking it's address when passing to wacom_get_report() is wrong. Signed-off-by: Ping Cheng Tested-by: Rafi Rubin Reviewed-by: Jason Gerecke Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/tablet/wacom_sys.c b/drivers/input/tablet/wacom_sys.c index b3a8bd3..23a933d 100644 --- a/drivers/input/tablet/wacom_sys.c +++ b/drivers/input/tablet/wacom_sys.c @@ -213,7 +213,7 @@ static void wacom_retrieve_report_data(struct usb_interface *intf, rep_data[0] = 12; result = wacom_get_report(intf, WAC_HID_FEATURE_REPORT, - rep_data[0], &rep_data, 2, + rep_data[0], rep_data, 2, WAC_MSG_RETRIES); if (result >= 0 && rep_data[1] > 2) -- cgit v0.10.2 From 1cecc5cc0658e128bcad0b29edb96f286066571d Mon Sep 17 00:00:00 2001 From: Ping Cheng Date: Thu, 28 Jun 2012 16:47:30 -0700 Subject: Input: wacom - don't retrieve touch_max when it is predefined Some models, such as 0xE6, report more fingers than we process. Reported-by: Jonathan Nieder Signed-off-by: Ping Cheng Tested-by: Nils Kanning Tested-by: Rafi Rubin Reviewed-by: Jason Gerecke Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/tablet/wacom_sys.c b/drivers/input/tablet/wacom_sys.c index 23a933d..b145841 100644 --- a/drivers/input/tablet/wacom_sys.c +++ b/drivers/input/tablet/wacom_sys.c @@ -398,7 +398,9 @@ static int wacom_parse_hid(struct usb_interface *intf, break; case HID_USAGE_CONTACTMAX: - wacom_retrieve_report_data(intf, features); + /* leave touch_max as is if predefined */ + if (!features->touch_max) + wacom_retrieve_report_data(intf, features); i++; break; } -- cgit v0.10.2 From 44b99462d9d776522e174d6c531ce5ccef309e26 Mon Sep 17 00:00:00 2001 From: Vaibhav Nagarnaik Date: Fri, 22 Jun 2012 11:50:05 -0700 Subject: ring-buffer: Fix crash due to uninitialized new_pages list head The new_pages list head in the cpu_buffer is not initialized. When adding pages to the ring buffer, if the memory allocation fails in ring_buffer_resize, the clean up handler tries to free up the allocated pages from all the cpu buffers. The panic is caused by referencing the uninitialized new_pages list head. Initializing the new_pages list head in rb_allocate_cpu_buffer fixes this. Link: http://lkml.kernel.org/r/1340391005-10880-1-git-send-email-vnagarnaik@google.com Cc: Justin Teravest Cc: David Sharp Signed-off-by: Vaibhav Nagarnaik Signed-off-by: Steven Rostedt diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 1d0f6a8..ba39cba 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -1075,6 +1075,7 @@ rb_allocate_cpu_buffer(struct ring_buffer *buffer, int nr_pages, int cpu) rb_init_page(bpage->page); INIT_LIST_HEAD(&cpu_buffer->reader_page->list); + INIT_LIST_HEAD(&cpu_buffer->new_pages); ret = rb_allocate_pages(cpu_buffer, nr_pages); if (ret < 0) -- cgit v0.10.2 From 48fdc72f23ad9a9956e524a47843135d0bbc3317 Mon Sep 17 00:00:00 2001 From: Vaibhav Nagarnaik Date: Fri, 29 Jun 2012 12:31:41 -0700 Subject: ring-buffer: Fix accounting of entries when removing pages When removing pages from the ring buffer, its state is not reset. This means that the counters need to be correctly updated to account for the pages removed. Update the overrun counter to reflect the removed events from the pages. Link: http://lkml.kernel.org/r/1340998301-1715-1-git-send-email-vnagarnaik@google.com Cc: Justin Teravest Cc: David Sharp Signed-off-by: Vaibhav Nagarnaik Signed-off-by: Steven Rostedt diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index ba39cba..f765465 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -1347,10 +1347,9 @@ rb_remove_pages(struct ring_buffer_per_cpu *cpu_buffer, unsigned int nr_pages) * If something was added to this page, it was full * since it is not the tail page. So we deduct the * bytes consumed in ring buffer from here. - * No need to update overruns, since this page is - * deleted from ring buffer and its entries are - * already accounted for. + * Increment overrun to account for the lost events. */ + local_add(page_entries, &cpu_buffer->overrun); local_sub(BUF_PAGE_SIZE, &cpu_buffer->entries_bytes); } -- cgit v0.10.2 From 6416c0409dda00aa66b2a4615044fb59d80f6bb2 Mon Sep 17 00:00:00 2001 From: Paul Parsons Date: Mon, 11 Jun 2012 15:31:14 +0100 Subject: ARM: pxa: hx4700: Fix basic suspend/resume Basic suspend/resume is fixed by ensuring that the PGSR registers are set correctly before sleep mode is entered. In particular four of the active low resets need to be driven high while in sleep mode, otherwise the unit resets itself instead of suspending. Another problem was that the PCFR_GPROD bit is set by the HTC bootloader; this caused GPIO reset (i.e. the reset button) to fail immediately after returning from sleep mode. Signed-off-by: Paul Parsons Cc: Philipp Zabel Signed-off-by: Haojian Zhuang diff --git a/arch/arm/mach-pxa/hx4700.c b/arch/arm/mach-pxa/hx4700.c index d09da6a..d3de84b 100644 --- a/arch/arm/mach-pxa/hx4700.c +++ b/arch/arm/mach-pxa/hx4700.c @@ -127,7 +127,11 @@ static unsigned long hx4700_pin_config[] __initdata = { GPIO19_SSP2_SCLK, GPIO86_SSP2_RXD, GPIO87_SSP2_TXD, - GPIO88_GPIO, + GPIO88_GPIO | MFP_LPM_DRIVE_HIGH, /* TSC2046_CS */ + + /* BQ24022 Regulator */ + GPIO72_GPIO | MFP_LPM_KEEP_OUTPUT, /* BQ24022_nCHARGE_EN */ + GPIO96_GPIO | MFP_LPM_KEEP_OUTPUT, /* BQ24022_ISET2 */ /* HX4700 specific input GPIOs */ GPIO12_GPIO | WAKEUP_ON_EDGE_RISE, /* ASIC3_IRQ */ @@ -135,6 +139,10 @@ static unsigned long hx4700_pin_config[] __initdata = { GPIO14_GPIO, /* nWLAN_IRQ */ /* HX4700 specific output GPIOs */ + GPIO61_GPIO | MFP_LPM_DRIVE_HIGH, /* W3220_nRESET */ + GPIO71_GPIO | MFP_LPM_DRIVE_HIGH, /* ASIC3_nRESET */ + GPIO81_GPIO | MFP_LPM_DRIVE_HIGH, /* CPU_GP_nRESET */ + GPIO116_GPIO | MFP_LPM_DRIVE_HIGH, /* CPU_HW_nRESET */ GPIO102_GPIO | MFP_LPM_DRIVE_LOW, /* SYNAPTICS_POWER_ON */ GPIO10_GPIO, /* GSM_IRQ */ @@ -872,14 +880,19 @@ static struct gpio global_gpios[] = { { GPIO110_HX4700_LCD_LVDD_3V3_ON, GPIOF_OUT_INIT_HIGH, "LCD_LVDD" }, { GPIO111_HX4700_LCD_AVDD_3V3_ON, GPIOF_OUT_INIT_HIGH, "LCD_AVDD" }, { GPIO32_HX4700_RS232_ON, GPIOF_OUT_INIT_HIGH, "RS232_ON" }, + { GPIO61_HX4700_W3220_nRESET, GPIOF_OUT_INIT_HIGH, "W3220_nRESET" }, { GPIO71_HX4700_ASIC3_nRESET, GPIOF_OUT_INIT_HIGH, "ASIC3_nRESET" }, + { GPIO81_HX4700_CPU_GP_nRESET, GPIOF_OUT_INIT_HIGH, "CPU_GP_nRESET" }, { GPIO82_HX4700_EUART_RESET, GPIOF_OUT_INIT_HIGH, "EUART_RESET" }, + { GPIO116_HX4700_CPU_HW_nRESET, GPIOF_OUT_INIT_HIGH, "CPU_HW_nRESET" }, }; static void __init hx4700_init(void) { int ret; + PCFR = PCFR_GPR_EN | PCFR_OPDE; + pxa2xx_mfp_config(ARRAY_AND_SIZE(hx4700_pin_config)); gpio_set_wake(GPIO12_HX4700_ASIC3_IRQ, 1); ret = gpio_request_array(ARRAY_AND_SIZE(global_gpios)); -- cgit v0.10.2 From 64941d8930e619ef37227f88c6ee17e2624c65d2 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 2 Jul 2012 15:06:22 +0900 Subject: sh: Fix up se7721 GPIOLIB=y build warnings. Presently the SH7720/21 serial code uses asm/gpio.h to get at the CPU GPIO port definitions, but in the case of GPIOLIB=y this also includes references to generic GPIOLIB routines that we don't have any function declarations for, tripping up on -Werror=implicit-function-declaration with newer gcc versions: CC arch/sh/kernel/cpu/sh3/serial-sh7720.o In file included from include/linux/sh_pfc.h:14:0, from arch/sh/include/asm/gpio.h:23, from arch/sh/kernel/cpu/sh3/serial-sh7720.c:5: include/asm-generic/gpio.h: In function 'gpio_get_value_cansleep': include/asm-generic/gpio.h:220:2: error: implicit declaration of function '__gpio_get_value' [-Werror=implicit-function-declaration] include/asm-generic/gpio.h: In function 'gpio_set_value_cansleep': include/asm-generic/gpio.h:226:2: error: implicit declaration of function '__gpio_set_value' [-Werror=implicit-function-declaration] In file included from arch/sh/include/asm/gpio.h:23:0, from arch/sh/kernel/cpu/sh3/serial-sh7720.c:5: include/linux/sh_pfc.h: At top level: include/linux/sh_pfc.h:121:19: error: field 'chip' has incomplete type Switch to using the cpu/ version for the port definitions explicitly. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh3/serial-sh7720.c b/arch/sh/kernel/cpu/sh3/serial-sh7720.c index 8832c52..c4a0336 100644 --- a/arch/sh/kernel/cpu/sh3/serial-sh7720.c +++ b/arch/sh/kernel/cpu/sh3/serial-sh7720.c @@ -2,7 +2,7 @@ #include #include #include -#include +#include static void sh7720_sci_init_pins(struct uart_port *port, unsigned int cflag) { -- cgit v0.10.2 From 396c89b327269c5a90bdd6152b5339e4cd2b8e73 Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Mon, 25 Jun 2012 01:17:02 -0300 Subject: ARM: imx27_visstrim_m10: Do not include commit 435ca24 (ARM i.MX: Visstrim_M10: Add board version detection) included , which is a header file about to be deleted according to 9f97da (Disintegrate asm/system.h for ARM) Include instead. Reported-by: Russell King Signed-off-by: Fabio Estevam Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c index f76edb9..ba09552 100644 --- a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c +++ b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c @@ -38,7 +38,7 @@ #include #include #include -#include +#include #include #include -- cgit v0.10.2 From 8f53dc724a83a0082184fa27df80c25c7df47340 Mon Sep 17 00:00:00 2001 From: Hiroshi DOYU Date: Wed, 27 Jun 2012 12:54:01 +0300 Subject: iommu/tegra: smmu: Fix unsleepable memory allocation allo_pdir() is called in smmu_iommu_domain_init() with spin_lock held. memory allocations in it have to be atomic/unsleepable. Signed-off-by: Hiroshi DOYU Reported-by: Chris Wright Acked-by: Chris Wright Cc: stable@vger.kernel.org Signed-off-by: Joerg Roedel diff --git a/drivers/iommu/tegra-smmu.c b/drivers/iommu/tegra-smmu.c index ecd6790..3f3d09d5 100644 --- a/drivers/iommu/tegra-smmu.c +++ b/drivers/iommu/tegra-smmu.c @@ -550,13 +550,13 @@ static int alloc_pdir(struct smmu_as *as) return 0; as->pte_count = devm_kzalloc(smmu->dev, - sizeof(as->pte_count[0]) * SMMU_PDIR_COUNT, GFP_KERNEL); + sizeof(as->pte_count[0]) * SMMU_PDIR_COUNT, GFP_ATOMIC); if (!as->pte_count) { dev_err(smmu->dev, "failed to allocate smmu_device PTE cunters\n"); return -ENOMEM; } - as->pdir_page = alloc_page(GFP_KERNEL | __GFP_DMA); + as->pdir_page = alloc_page(GFP_ATOMIC | __GFP_DMA); if (!as->pdir_page) { dev_err(smmu->dev, "failed to allocate smmu_device page directory\n"); -- cgit v0.10.2 From 68ee6d22376411f8ec668413f1b632a34192a807 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 27 Jun 2012 12:08:55 +0300 Subject: dma-debug: debugfs_create_bool() takes a u32 pointer Even though it has "bool" in the name, you have pass a u32 pointer to debugfs_create_bool(). Otherwise you get memory corruption in write_file_bool(). Fortunately in this case the corruption happens in an alignment hole between variables so it doesn't cause any problems. Signed-off-by: Dan Carpenter Signed-off-by: Joerg Roedel diff --git a/lib/dma-debug.c b/lib/dma-debug.c index 518aea7..66ce414 100644 --- a/lib/dma-debug.c +++ b/lib/dma-debug.c @@ -78,7 +78,7 @@ static LIST_HEAD(free_entries); static DEFINE_SPINLOCK(free_entries_lock); /* Global disable flag - will be set in case of an error */ -static bool global_disable __read_mostly; +static u32 global_disable __read_mostly; /* Global error count */ static u32 error_count; @@ -657,7 +657,7 @@ static int dma_debug_fs_init(void) global_disable_dent = debugfs_create_bool("disabled", 0444, dma_debug_dent, - (u32 *)&global_disable); + &global_disable); if (!global_disable_dent) goto out_err; -- cgit v0.10.2 From 3775d4818d72081e2afa2aed2442a2b9ecfc5eab Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 27 Jun 2012 12:09:18 +0300 Subject: iommu/amd: fix type bug in flush code write_file_bool() modifies 32 bits of data, so "amd_iommu_unmap_flush" needs to be 32 bits as well or we'll corrupt memory. Fortunately it looks like the data is aligned with a gap after the declaration so this is harmless in production. Signed-off-by: Dan Carpenter Signed-off-by: Joerg Roedel diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c index dfe7d37..6256263 100644 --- a/drivers/iommu/amd_iommu.c +++ b/drivers/iommu/amd_iommu.c @@ -404,7 +404,7 @@ static void amd_iommu_stats_init(void) return; de_fflush = debugfs_create_bool("fullflush", 0444, stats_dir, - (u32 *)&amd_iommu_unmap_flush); + &amd_iommu_unmap_flush); amd_iommu_stats_add(&compl_wait); amd_iommu_stats_add(&cnt_map_single); diff --git a/drivers/iommu/amd_iommu_init.c b/drivers/iommu/amd_iommu_init.c index c04ddca..a33612f 100644 --- a/drivers/iommu/amd_iommu_init.c +++ b/drivers/iommu/amd_iommu_init.c @@ -129,7 +129,7 @@ u16 amd_iommu_last_bdf; /* largest PCI device id we have to handle */ LIST_HEAD(amd_iommu_unity_map); /* a list of required unity mappings we find in ACPI */ -bool amd_iommu_unmap_flush; /* if true, flush on every unmap */ +u32 amd_iommu_unmap_flush; /* if true, flush on every unmap */ LIST_HEAD(amd_iommu_list); /* list of all AMD IOMMUs in the system */ diff --git a/drivers/iommu/amd_iommu_types.h b/drivers/iommu/amd_iommu_types.h index 2435555..c1b1d48 100644 --- a/drivers/iommu/amd_iommu_types.h +++ b/drivers/iommu/amd_iommu_types.h @@ -652,7 +652,7 @@ extern unsigned long *amd_iommu_pd_alloc_bitmap; * If true, the addresses will be flushed on unmap time, not when * they are reused */ -extern bool amd_iommu_unmap_flush; +extern u32 amd_iommu_unmap_flush; /* Smallest number of PASIDs supported by any IOMMU in the system */ extern u32 amd_iommu_max_pasids; -- cgit v0.10.2 From 95669d788192d351890bfb16785336c2db816356 Mon Sep 17 00:00:00 2001 From: Kevin Hilman Date: Thu, 28 Jun 2012 10:01:31 -0700 Subject: ARM: OMAP2: Overo: init I2C before MMC to fix MMC suspend/resume failure In order for suspend/resume dependencies to work correctly, I2C has to be initialized (more specifically, registered with the driver core) before MMC. Without this, the MMC driver fails to adjust the VMMC regulator (using i2c writes) during the suspend path. Problem found testing suspend/resume on 3730/OveroSTORM platform. Signed-off-by: Kevin Hilman Signed-off-by: Tony Lindgren diff --git a/arch/arm/mach-omap2/board-overo.c b/arch/arm/mach-omap2/board-overo.c index 8fa2fc3..779734d 100644 --- a/arch/arm/mach-omap2/board-overo.c +++ b/arch/arm/mach-omap2/board-overo.c @@ -494,8 +494,8 @@ static void __init overo_init(void) regulator_register_fixed(0, dummy_supplies, ARRAY_SIZE(dummy_supplies)); omap3_mux_init(board_mux, OMAP_PACKAGE_CBB); - omap_hsmmc_init(mmc); overo_i2c_init(); + omap_hsmmc_init(mmc); omap_display_init(&overo_dss_data); omap_serial_init(); omap_sdrc_init(mt46h32m32lf6_sdrc_params, -- cgit v0.10.2 From 5941b8142e9f4d7b8b99562e833067b307b3f390 Mon Sep 17 00:00:00 2001 From: Kevin Hilman Date: Thu, 28 Jun 2012 10:01:32 -0700 Subject: ARM: OMAP4: TWL6030: ensure sys_nirq1 is mux'd and wakeup enabled The SYS_NIRQ1 pin is the interupt line for the PMIC part of the TWL6030 and interrupts from the PMIC are needed as wakeup sources. Ensure this pin is mux'd as input and has wakeup enabled so PMIC interupts (e.g. RTC) can be used as wakeup sources. Tested on OMAP4430/Panda. Signed-off-by: Kevin Hilman Signed-off-by: Tony Lindgren diff --git a/arch/arm/mach-omap2/twl-common.c b/arch/arm/mach-omap2/twl-common.c index 119d5a9..43a9790 100644 --- a/arch/arm/mach-omap2/twl-common.c +++ b/arch/arm/mach-omap2/twl-common.c @@ -32,6 +32,7 @@ #include "twl-common.h" #include "pm.h" #include "voltage.h" +#include "mux.h" static struct i2c_board_info __initdata pmic_i2c_board_info = { .addr = 0x48, @@ -77,6 +78,7 @@ void __init omap4_pmic_init(const char *pmic_type, struct twl6040_platform_data *twl6040_data, int twl6040_irq) { /* PMIC part*/ + omap_mux_init_signal("sys_nirq1", OMAP_PIN_INPUT_PULLUP | OMAP_PIN_OFF_WAKEUPENABLE); strncpy(omap4_i2c1_board_info[0].type, pmic_type, sizeof(omap4_i2c1_board_info[0].type)); omap4_i2c1_board_info[0].irq = OMAP44XX_IRQ_SYS_1N; -- cgit v0.10.2 From 76a8349dfdb775d387e9767db3092e410403138a Mon Sep 17 00:00:00 2001 From: David Ahern Date: Thu, 14 Jun 2012 12:36:17 -0600 Subject: perf script: Fix format regression due to libtraceevent merge Consider the commands: perf record -e sched:sched_switch -fo /tmp/perf.data -a -- sleep 1 perf script -i /tmp/perf.data In v3.4 the output has the form (lines wrapped here) perf 29214 [005] 821043.582596: sched_switch: prev_comm=perf prev_pid=29214 prev_prio=120 prev_state=S ==> next_comm=swapper/5 next_pid=0 next_prio=120 In 3.5 that same line has become: perf 29214 [005] 821043.582596: sched_switch: <...>-29214 [005] 0.000000000: sched_switch: prev_comm=perf prev_pid=29214 prev_prio=120 prev_state=S ==> next_comm=swapper/5 next_pid=0 next_prio=120 Note the duplicates in the output -- pid, cpu, event name. With this patch the v3.4 output is restored: perf 29214 [005] 821043.582596: sched_switch: prev_comm=perf prev_pid=29214 prev_prio=120 prev_state=S ==> next_comm=swapper/5 next_pid=0 next_prio=120 v3: Remove that pesky newline too. Output now matches v3.4 (pre-libtracevent). v2: Change print_trace_event function local to perf per Steve's comments. Signed-off-by: David Ahern Acked-by: Steven Rostedt Cc: Frederic Weisbecker Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Steven Rostedt Link: http://lkml.kernel.org/r/1339698977-68962-1-git-send-email-dsahern@gmail.com Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c index df2fddb..5dd3b5e 100644 --- a/tools/perf/util/trace-event-parse.c +++ b/tools/perf/util/trace-event-parse.c @@ -198,9 +198,8 @@ void print_trace_event(int cpu, void *data, int size) record.data = data; trace_seq_init(&s); - pevent_print_event(pevent, &s, &record); + pevent_event_info(&s, event, &record); trace_seq_do_printf(&s); - printf("\n"); } void print_event(int cpu, void *data, int size, unsigned long long nsecs, -- cgit v0.10.2 From 207b5792696206663a38e525b9793644895bad3b Mon Sep 17 00:00:00 2001 From: David Ahern Date: Sun, 1 Jul 2012 16:11:37 -0600 Subject: perf kvm: Fix regression with guest machine creation Commit 743eb868657bdb1b26c7b24077ca21c67c82c777 reworked when the machines were created. Prior to this commit guest machines could be created in perf_event__process_kernel_mmap() while processing kernel MMAP events. This commit assumes that the machines exist by the time perf_session_deliver_event is called (e.g., during processing of build id events) - which is not always correct. One example is the use of default guest args (--guestkallsyms and --guestmodules) for short times where no samples hit within a guest module. For this case no build id is added to the file header. No build id == no machine created. That leads to the next example -- the use of no-buildid (-B) on the record for all perf-kvm invocations. In both cases perf report dies with a SEGFAULT of the form: (gdb) bt 0 0x000000000046dd7b in machine__mmap_name (self=0x0, bf=0x7fffffffbd20 "q\021", size=4096) at util/map.c:715 1 0x0000000000444161 in perf_event__process_kernel_mmap (tool=0x7fffffffdd80, event=0x7ffff7fb4120, machine=0x0) at util/event.c:562 2 0x0000000000444642 in perf_event__process_mmap (tool=0x7fffffffdd80, event=0x7ffff7fb4120, sample=0x7fffffffd210, machine=0x0) at util/event.c:668 3 0x0000000000470e0b in perf_session_deliver_event (session=0x915ca0, event=0x7ffff7fb4120, sample=0x7fffffffd210, tool=0x7fffffffdd80, file_offset=8480) at util/session.c:979 4 0x000000000047032e in flush_sample_queue (s=0x915ca0, tool=0x7fffffffdd80) at util/session.c:679 5 0x0000000000471c8d in __perf_session__process_events (session=0x915ca0, data_offset=400, data_size=150448, file_size=150848, tool= 0x7fffffffdd80) at util/session.c:1363 6 0x0000000000471d42 in perf_session__process_events (self=0x915ca0, tool=0x7fffffffdd80) at util/session.c:1379 7 0x000000000042484a in __cmd_report (rep=0x7fffffffdd80) at builtin-report.c:368 8 0x0000000000425bf1 in cmd_report (argc=0, argv=0x915b00, prefix=0x0) at builtin-report.c:756 9 0x0000000000438505 in __cmd_report (argc=4, argv=0x7fffffffe260) at builtin-kvm.c:84 10 0x000000000043882a in cmd_kvm (argc=4, argv=0x7fffffffe260, prefix=0x0) at builtin-kvm.c:131 11 0x00000000004152cd in run_builtin (p=0x7a54e8, argc=9, argv=0x7fffffffe260) at perf.c:273 12 0x00000000004154c7 in handle_internal_command (argc=9, argv=0x7fffffffe260) at perf.c:345 13 0x0000000000415613 in run_argv (argcp=0x7fffffffe14c, argv=0x7fffffffe140) at perf.c:389 14 0x0000000000415899 in main (argc=9, argv=0x7fffffffe260) at perf.c:487 Fix by allowing the machine to be created in perf_session_deliver_event. Tested with --guestmount option and default guest args, with and without -B arg on record for both and for short (10 seconds) and long (10 minutes) windows. Reported-by: Pradeep Kumar Surisetty Signed-off-by: David Ahern Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Jiri Olsa Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Pradeep Kumar Surisetty Link: http://lkml.kernel.org/r/1341180697-64515-1-git-send-email-dsahern@gmail.com Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index c3e399b..56142d0 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -926,7 +926,7 @@ static struct machine * else pid = event->ip.pid; - return perf_session__find_machine(session, pid); + return perf_session__findnew_machine(session, pid); } return perf_session__find_host_machine(session); -- cgit v0.10.2 From 7ed97ad41ffa94040dfd593948962a7e9e7b0db9 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Mon, 2 Jul 2012 09:12:57 -0600 Subject: perf kvm: Fix segfault with report and mixed guestmount use Using the guestmount option on record: $ perf kvm --guest --host --guestmount=/tmp/guest-mount record -ag But not the subsequent report: $ perf kvm report causes a SEGFAULT in the usual place: (gdb) bt 0 0x0000000000470356 in machine__mmap_name (self=0x0, bf=0x7fffffffbdb0 " z\370\367\377\177", size= 4096) at util/map.c:712 1 0x00000000004453e8 in perf_event__process_kernel_mmap (tool=0x7fffffffde10, event=0x7ffff7f87e38, machine=0x0) at util/event.c:550 2 0x00000000004458c9 in perf_event__process_mmap (tool=0x7fffffffde10, event=0x7ffff7f87e38, sample= 0x7fffffffd2a0, machine=0x0) at util/event.c:656 3 0x00000000004733e0 in perf_session_deliver_event (session=0x91aca0, event=0x7ffff7f87e38, sample= 0x7fffffffd2a0, tool=0x7fffffffde10, file_offset=7736) at util/session.c:979 ... The MMAP events in this case already contain the full path to the module. No need to require it for the report path to. Signed-off-by: David Ahern Cc: Frederic Weisbecker Cc: Ingo Molnar Cc: Jiri Olsa Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/1341241977-71535-1-git-send-email-dsahern@gmail.com Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index 35ae568..a1f4e36 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -669,25 +669,26 @@ struct machine *machines__find(struct rb_root *self, pid_t pid) struct machine *machines__findnew(struct rb_root *self, pid_t pid) { char path[PATH_MAX]; - const char *root_dir; + const char *root_dir = ""; struct machine *machine = machines__find(self, pid); - if (!machine || machine->pid != pid) { - if (pid == HOST_KERNEL_ID || pid == DEFAULT_GUEST_KERNEL_ID) - root_dir = ""; - else { - if (!symbol_conf.guestmount) - goto out; - sprintf(path, "%s/%d", symbol_conf.guestmount, pid); - if (access(path, R_OK)) { - pr_err("Can't access file %s\n", path); - goto out; - } - root_dir = path; + if (machine && (machine->pid == pid)) + goto out; + + if ((pid != HOST_KERNEL_ID) && + (pid != DEFAULT_GUEST_KERNEL_ID) && + (symbol_conf.guestmount)) { + sprintf(path, "%s/%d", symbol_conf.guestmount, pid); + if (access(path, R_OK)) { + pr_err("Can't access file %s\n", path); + machine = NULL; + goto out; } - machine = machines__add(self, pid, root_dir); + root_dir = path; } + machine = machines__add(self, pid, root_dir); + out: return machine; } -- cgit v0.10.2 From cba6d0d64ee53772b285d0c0c288deefbeaf7775 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Mon, 2 Jul 2012 07:08:42 -0700 Subject: Revert "rcu: Move PREEMPT_RCU preemption to switch_to() invocation" This reverts commit 616c310e83b872024271c915c1b9ab505b9efad9. (Move PREEMPT_RCU preemption to switch_to() invocation). Testing by Sasha Levin showed that this can result in deadlock due to invoking the scheduler when one of the runqueue locks is held. Because this commit was simply a performance optimization, revert it. Reported-by: Sasha Levin Signed-off-by: Paul E. McKenney Tested-by: Sasha Levin diff --git a/arch/um/drivers/mconsole_kern.c b/arch/um/drivers/mconsole_kern.c index 88e466b..43b39d6 100644 --- a/arch/um/drivers/mconsole_kern.c +++ b/arch/um/drivers/mconsole_kern.c @@ -705,7 +705,6 @@ static void stack_proc(void *arg) struct task_struct *from = current, *to = arg; to->thread.saved_task = from; - rcu_switch_from(from); switch_to(from, to, from); } diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 26d1a47..9cac722 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -184,7 +184,6 @@ static inline int rcu_preempt_depth(void) /* Internal to kernel */ extern void rcu_sched_qs(int cpu); extern void rcu_bh_qs(int cpu); -extern void rcu_preempt_note_context_switch(void); extern void rcu_check_callbacks(int cpu, int user); struct notifier_block; extern void rcu_idle_enter(void); diff --git a/include/linux/rcutiny.h b/include/linux/rcutiny.h index 854dc4c..4e56a9c 100644 --- a/include/linux/rcutiny.h +++ b/include/linux/rcutiny.h @@ -87,6 +87,10 @@ static inline void kfree_call_rcu(struct rcu_head *head, #ifdef CONFIG_TINY_RCU +static inline void rcu_preempt_note_context_switch(void) +{ +} + static inline int rcu_needs_cpu(int cpu, unsigned long *delta_jiffies) { *delta_jiffies = ULONG_MAX; @@ -95,6 +99,7 @@ static inline int rcu_needs_cpu(int cpu, unsigned long *delta_jiffies) #else /* #ifdef CONFIG_TINY_RCU */ +void rcu_preempt_note_context_switch(void); int rcu_preempt_needs_cpu(void); static inline int rcu_needs_cpu(int cpu, unsigned long *delta_jiffies) @@ -108,6 +113,7 @@ static inline int rcu_needs_cpu(int cpu, unsigned long *delta_jiffies) static inline void rcu_note_context_switch(int cpu) { rcu_sched_qs(cpu); + rcu_preempt_note_context_switch(); } /* diff --git a/include/linux/sched.h b/include/linux/sched.h index 4059c0f..06a4c5f 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1871,22 +1871,12 @@ static inline void rcu_copy_process(struct task_struct *p) INIT_LIST_HEAD(&p->rcu_node_entry); } -static inline void rcu_switch_from(struct task_struct *prev) -{ - if (prev->rcu_read_lock_nesting != 0) - rcu_preempt_note_context_switch(); -} - #else static inline void rcu_copy_process(struct task_struct *p) { } -static inline void rcu_switch_from(struct task_struct *prev) -{ -} - #endif #ifdef CONFIG_SMP diff --git a/kernel/rcutree.c b/kernel/rcutree.c index 38ecdda..4b97bba 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -201,6 +201,7 @@ void rcu_note_context_switch(int cpu) { trace_rcu_utilization("Start context switch"); rcu_sched_qs(cpu); + rcu_preempt_note_context_switch(cpu); trace_rcu_utilization("End context switch"); } EXPORT_SYMBOL_GPL(rcu_note_context_switch); diff --git a/kernel/rcutree.h b/kernel/rcutree.h index ea05649..19b61ac 100644 --- a/kernel/rcutree.h +++ b/kernel/rcutree.h @@ -444,6 +444,7 @@ DECLARE_PER_CPU(char, rcu_cpu_has_work); /* Forward declarations for rcutree_plugin.h */ static void rcu_bootup_announce(void); long rcu_batches_completed(void); +static void rcu_preempt_note_context_switch(int cpu); static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp); #ifdef CONFIG_HOTPLUG_CPU static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp, diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h index 5271a02..3e48994 100644 --- a/kernel/rcutree_plugin.h +++ b/kernel/rcutree_plugin.h @@ -153,7 +153,7 @@ static void rcu_preempt_qs(int cpu) * * Caller must disable preemption. */ -void rcu_preempt_note_context_switch(void) +static void rcu_preempt_note_context_switch(int cpu) { struct task_struct *t = current; unsigned long flags; @@ -164,7 +164,7 @@ void rcu_preempt_note_context_switch(void) (t->rcu_read_unlock_special & RCU_READ_UNLOCK_BLOCKED) == 0) { /* Possibly blocking in an RCU read-side critical section. */ - rdp = __this_cpu_ptr(rcu_preempt_state.rda); + rdp = per_cpu_ptr(rcu_preempt_state.rda, cpu); rnp = rdp->mynode; raw_spin_lock_irqsave(&rnp->lock, flags); t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BLOCKED; @@ -228,7 +228,7 @@ void rcu_preempt_note_context_switch(void) * means that we continue to block the current grace period. */ local_irq_save(flags); - rcu_preempt_qs(smp_processor_id()); + rcu_preempt_qs(cpu); local_irq_restore(flags); } @@ -1002,6 +1002,14 @@ void rcu_force_quiescent_state(void) EXPORT_SYMBOL_GPL(rcu_force_quiescent_state); /* + * Because preemptible RCU does not exist, we never have to check for + * CPUs being in quiescent states. + */ +static void rcu_preempt_note_context_switch(int cpu) +{ +} + +/* * Because preemptible RCU does not exist, there are never any preempted * RCU readers. */ diff --git a/kernel/sched/core.c b/kernel/sched/core.c index d5594a4..eaead2d 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2081,7 +2081,6 @@ context_switch(struct rq *rq, struct task_struct *prev, #endif /* Here we just switch the register state and the stack. */ - rcu_switch_from(prev); switch_to(prev, next, prev); barrier(); -- cgit v0.10.2 From 597a60fadedf9a40fdff8735054bf772b3dafd57 Mon Sep 17 00:00:00 2001 From: Stefan Behrens Date: Thu, 14 Jun 2012 16:42:31 +0200 Subject: Btrfs: don't count I/O statistic read errors for missing devices It is normal behaviour of the low level btrfs function btrfs_map_bio() to complete a bio with -EIO if the device is missing, instead of just preventing the bio creation in an earlier step. This used to cause I/O statistic read error increments and annoying printk_ratelimited messages. This commit fixes the issue. Signed-off-by: Stefan Behrens Reported-by: Carey Underwood diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 8a3d259..3f292cf 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -4061,16 +4061,18 @@ static void btrfs_end_bio(struct bio *bio, int err) BUG_ON(stripe_index >= bbio->num_stripes); dev = bbio->stripes[stripe_index].dev; - if (bio->bi_rw & WRITE) - btrfs_dev_stat_inc(dev, - BTRFS_DEV_STAT_WRITE_ERRS); - else - btrfs_dev_stat_inc(dev, - BTRFS_DEV_STAT_READ_ERRS); - if ((bio->bi_rw & WRITE_FLUSH) == WRITE_FLUSH) - btrfs_dev_stat_inc(dev, - BTRFS_DEV_STAT_FLUSH_ERRS); - btrfs_dev_stat_print_on_error(dev); + if (dev->bdev) { + if (bio->bi_rw & WRITE) + btrfs_dev_stat_inc(dev, + BTRFS_DEV_STAT_WRITE_ERRS); + else + btrfs_dev_stat_inc(dev, + BTRFS_DEV_STAT_READ_ERRS); + if ((bio->bi_rw & WRITE_FLUSH) == WRITE_FLUSH) + btrfs_dev_stat_inc(dev, + BTRFS_DEV_STAT_FLUSH_ERRS); + btrfs_dev_stat_print_on_error(dev); + } } } -- cgit v0.10.2 From c3473e830074ef04f974f2829690942dd8580619 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 19 Jun 2012 10:59:00 -0400 Subject: Btrfs: fix dio write vs buffered read race Miao pointed out there's a problem with mixing dio writes and buffered reads. If the read happens between us invalidating the page range and actually locking the extent we can bring in pages into page cache. Then once the write finishes if somebody tries to read again it will just find uptodate pages and we'll read stale data. So we need to lock the extent and check for uptodate bits in the range. If there are uptodate bits we need to unlock and invalidate again. This will keep this race from happening since we will hold the extent locked until we create the ordered extent, and then teh read side always waits for ordered extents. There was also a race in how we updated i_size, previously we were relying on the generic DIO stuff to adjust the i_size after the DIO had completed, but this happens outside of the extent lock which means reads could come in and not see the updated i_size. So instead move this work into where we create the extents, and then this way the update ordered i_size stuff works properly in the endio handlers. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 876cddd..248d202 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1334,7 +1334,6 @@ static ssize_t __btrfs_direct_write(struct kiocb *iocb, loff_t *ppos, size_t count, size_t ocount) { struct file *file = iocb->ki_filp; - struct inode *inode = fdentry(file)->d_inode; struct iov_iter i; ssize_t written; ssize_t written_buffered; @@ -1344,18 +1343,6 @@ static ssize_t __btrfs_direct_write(struct kiocb *iocb, written = generic_file_direct_write(iocb, iov, &nr_segs, pos, ppos, count, ocount); - /* - * the generic O_DIRECT will update in-memory i_size after the - * DIOs are done. But our endio handlers that update the on - * disk i_size never update past the in memory i_size. So we - * need one more update here to catch any additions to the - * file - */ - if (inode->i_size != BTRFS_I(inode)->disk_i_size) { - btrfs_ordered_update_i_size(inode, inode->i_size, NULL); - mark_inode_dirty(inode); - } - if (written < 0 || written == count) return written; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 4a4f2d5..6971fb5 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5904,8 +5904,17 @@ map: bh_result->b_size = len; bh_result->b_bdev = em->bdev; set_buffer_mapped(bh_result); - if (create && !test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) - set_buffer_new(bh_result); + if (create) { + if (!test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) + set_buffer_new(bh_result); + + /* + * Need to update the i_size under the extent lock so buffered + * readers will get the updated i_size when we unlock. + */ + if (start + len > i_size_read(inode)) + i_size_write(inode, start + len); + } free_extent_map(em); @@ -6388,12 +6397,48 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, */ ordered = btrfs_lookup_ordered_range(inode, lockstart, lockend - lockstart + 1); - if (!ordered) + + /* + * We need to make sure there are no buffered pages in this + * range either, we could have raced between the invalidate in + * generic_file_direct_write and locking the extent. The + * invalidate needs to happen so that reads after a write do not + * get stale data. + */ + if (!ordered && (!writing || + !test_range_bit(&BTRFS_I(inode)->io_tree, + lockstart, lockend, EXTENT_UPTODATE, 0, + cached_state))) break; + unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend, &cached_state, GFP_NOFS); - btrfs_start_ordered_extent(inode, ordered, 1); - btrfs_put_ordered_extent(ordered); + + if (ordered) { + btrfs_start_ordered_extent(inode, ordered, 1); + btrfs_put_ordered_extent(ordered); + } else { + /* Screw you mmap */ + ret = filemap_write_and_wait_range(file->f_mapping, + lockstart, + lockend); + if (ret) + goto out; + + /* + * If we found a page that couldn't be invalidated just + * fall back to buffered. + */ + ret = invalidate_inode_pages2_range(file->f_mapping, + lockstart >> PAGE_CACHE_SHIFT, + lockend >> PAGE_CACHE_SHIFT); + if (ret) { + if (ret == -EBUSY) + ret = 0; + goto out; + } + } + cond_resched(); } -- cgit v0.10.2 From 68310a5e42f93c2242ec1836c3b18d531e0065e2 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Fri, 22 Jun 2012 12:24:12 -0600 Subject: Btrfs: restore restriper state on all mounts Fix a bug that triggered asserts in btrfs_balance() in both normal and resume modes -- restriper state was not properly restored on read-only mounts. This factors out resuming code from btrfs_restore_balance(), which is now also called earlier in the mount sequence to avoid the problem of some early writes getting the old profile. Signed-off-by: Ilya Dryomov diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 7d7bc8e..3a7961b 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2354,12 +2354,17 @@ retry_root_backup: BTRFS_CSUM_TREE_OBJECTID, csum_root); if (ret) goto recovery_tree_root; - csum_root->track_dirty = 1; fs_info->generation = generation; fs_info->last_trans_committed = generation; + ret = btrfs_recover_balance(fs_info); + if (ret) { + printk(KERN_WARNING "btrfs: failed to recover balance\n"); + goto fail_block_groups; + } + ret = btrfs_init_dev_stats(fs_info); if (ret) { printk(KERN_ERR "btrfs: failed to init dev_stats: %d\n", @@ -2492,9 +2497,6 @@ retry_root_backup: err = btrfs_orphan_cleanup(fs_info->tree_root); up_read(&fs_info->cleanup_work_sem); - if (!err) - err = btrfs_recover_balance(fs_info->tree_root); - if (err) { close_ctree(tree_root); return err; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 3f292cf..48943d0 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2867,9 +2867,8 @@ static int balance_kthread(void *data) return ret; } -int btrfs_recover_balance(struct btrfs_root *tree_root) +int btrfs_recover_balance(struct btrfs_fs_info *fs_info) { - struct task_struct *tsk; struct btrfs_balance_control *bctl; struct btrfs_balance_item *item; struct btrfs_disk_balance_args disk_bargs; @@ -2882,29 +2881,30 @@ int btrfs_recover_balance(struct btrfs_root *tree_root) if (!path) return -ENOMEM; - bctl = kzalloc(sizeof(*bctl), GFP_NOFS); - if (!bctl) { - ret = -ENOMEM; - goto out; - } - key.objectid = BTRFS_BALANCE_OBJECTID; key.type = BTRFS_BALANCE_ITEM_KEY; key.offset = 0; - ret = btrfs_search_slot(NULL, tree_root, &key, path, 0, 0); + ret = btrfs_search_slot(NULL, fs_info->tree_root, &key, path, 0, 0); if (ret < 0) - goto out_bctl; + goto out; if (ret > 0) { /* ret = -ENOENT; */ ret = 0; - goto out_bctl; + goto out; + } + + bctl = kzalloc(sizeof(*bctl), GFP_NOFS); + if (!bctl) { + ret = -ENOMEM; + goto out; } leaf = path->nodes[0]; item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_balance_item); - bctl->fs_info = tree_root->fs_info; - bctl->flags = btrfs_balance_flags(leaf, item) | BTRFS_BALANCE_RESUME; + bctl->fs_info = fs_info; + bctl->flags = btrfs_balance_flags(leaf, item); + bctl->flags |= BTRFS_BALANCE_RESUME; btrfs_balance_data(leaf, item, &disk_bargs); btrfs_disk_balance_args_to_cpu(&bctl->data, &disk_bargs); @@ -2913,14 +2913,13 @@ int btrfs_recover_balance(struct btrfs_root *tree_root) btrfs_balance_sys(leaf, item, &disk_bargs); btrfs_disk_balance_args_to_cpu(&bctl->sys, &disk_bargs); - tsk = kthread_run(balance_kthread, bctl, "btrfs-balance"); - if (IS_ERR(tsk)) - ret = PTR_ERR(tsk); - else - goto out; + mutex_lock(&fs_info->volume_mutex); + mutex_lock(&fs_info->balance_mutex); -out_bctl: - kfree(bctl); + set_balance_control(bctl); + + mutex_unlock(&fs_info->balance_mutex); + mutex_unlock(&fs_info->volume_mutex); out: btrfs_free_path(path); return ret; diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 74366f2..e1b1a64 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -281,7 +281,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs); -int btrfs_recover_balance(struct btrfs_root *tree_root); +int btrfs_recover_balance(struct btrfs_fs_info *fs_info); int btrfs_pause_balance(struct btrfs_fs_info *fs_info); int btrfs_cancel_balance(struct btrfs_fs_info *fs_info); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); -- cgit v0.10.2 From 2b6ba629b5aac51e7099efbb43e2b403213aa7fb Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Fri, 22 Jun 2012 12:24:13 -0600 Subject: Btrfs: resume balance on rw (re)mounts properly This introduces btrfs_resume_balance_async(), which, given that restriper state was recovered earlier by btrfs_recover_balance(), resumes balance in btrfs-balance kthread. Signed-off-by: Ilya Dryomov diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 3a7961b..8cc4710 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2490,17 +2490,23 @@ retry_root_backup: goto fail_trans_kthread; } - if (!(sb->s_flags & MS_RDONLY)) { - down_read(&fs_info->cleanup_work_sem); - err = btrfs_orphan_cleanup(fs_info->fs_root); - if (!err) - err = btrfs_orphan_cleanup(fs_info->tree_root); + if (sb->s_flags & MS_RDONLY) + return 0; + + down_read(&fs_info->cleanup_work_sem); + if ((ret = btrfs_orphan_cleanup(fs_info->fs_root)) || + (ret = btrfs_orphan_cleanup(fs_info->tree_root))) { up_read(&fs_info->cleanup_work_sem); + close_ctree(tree_root); + return ret; + } + up_read(&fs_info->cleanup_work_sem); - if (err) { - close_ctree(tree_root); - return err; - } + ret = btrfs_resume_balance_async(fs_info); + if (ret) { + printk(KERN_WARNING "btrfs: failed to resume balance\n"); + close_ctree(tree_root); + return ret; } return 0; diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 0eb9a4d..e239915 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1187,6 +1187,10 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) if (ret) goto restore; + ret = btrfs_resume_balance_async(fs_info); + if (ret) + goto restore; + sb->s_flags &= ~MS_RDONLY; } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 48943d0..ecaad40 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2845,28 +2845,46 @@ out: static int balance_kthread(void *data) { - struct btrfs_balance_control *bctl = - (struct btrfs_balance_control *)data; - struct btrfs_fs_info *fs_info = bctl->fs_info; + struct btrfs_fs_info *fs_info = data; int ret = 0; mutex_lock(&fs_info->volume_mutex); mutex_lock(&fs_info->balance_mutex); - set_balance_control(bctl); - - if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) { - printk(KERN_INFO "btrfs: force skipping balance\n"); - } else { + if (fs_info->balance_ctl) { printk(KERN_INFO "btrfs: continuing balance\n"); - ret = btrfs_balance(bctl, NULL); + ret = btrfs_balance(fs_info->balance_ctl, NULL); } mutex_unlock(&fs_info->balance_mutex); mutex_unlock(&fs_info->volume_mutex); + return ret; } +int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info) +{ + struct task_struct *tsk; + + spin_lock(&fs_info->balance_lock); + if (!fs_info->balance_ctl) { + spin_unlock(&fs_info->balance_lock); + return 0; + } + spin_unlock(&fs_info->balance_lock); + + if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) { + printk(KERN_INFO "btrfs: force skipping balance\n"); + return 0; + } + + tsk = kthread_run(balance_kthread, fs_info, "btrfs-balance"); + if (IS_ERR(tsk)) + return PTR_ERR(tsk); + + return 0; +} + int btrfs_recover_balance(struct btrfs_fs_info *fs_info) { struct btrfs_balance_control *bctl; diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index e1b1a64..95f6637 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -281,6 +281,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs); +int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info); int btrfs_recover_balance(struct btrfs_fs_info *fs_info); int btrfs_pause_balance(struct btrfs_fs_info *fs_info); int btrfs_cancel_balance(struct btrfs_fs_info *fs_info); -- cgit v0.10.2 From d3a94048c912e18e99a091d5ea2d0a1178152d6f Mon Sep 17 00:00:00 2001 From: Alexander Block Date: Mon, 25 Jun 2012 15:36:12 -0600 Subject: Btrfs: use _IOR for BTRFS_IOC_SUBVOL_GETFLAGS We used the wrong ioctl macro for the getflags ioctl before. As we don't have the set/getflags ioctls in the user space ioctl.h at the moment, it's safe to fix it now. Reviewed-by: David Sterba Signed-off-by: Alexander Block diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index 497c530..e440aa6 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -339,7 +339,7 @@ struct btrfs_ioctl_get_dev_stats { #define BTRFS_IOC_WAIT_SYNC _IOW(BTRFS_IOCTL_MAGIC, 22, __u64) #define BTRFS_IOC_SNAP_CREATE_V2 _IOW(BTRFS_IOCTL_MAGIC, 23, \ struct btrfs_ioctl_vol_args_v2) -#define BTRFS_IOC_SUBVOL_GETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 25, __u64) +#define BTRFS_IOC_SUBVOL_GETFLAGS _IOR(BTRFS_IOCTL_MAGIC, 25, __u64) #define BTRFS_IOC_SUBVOL_SETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 26, __u64) #define BTRFS_IOC_SCRUB _IOWR(BTRFS_IOCTL_MAGIC, 27, \ struct btrfs_ioctl_scrub_args) -- cgit v0.10.2 From 6bf02314d9a5c29f6ec30285b9ad5361c2d4c85a Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 25 Jun 2012 21:59:09 -0600 Subject: Btrfs: fix wrong check during log recovery When we're evicting an inode during log recovery, we need to ensure that the inode is not in orphan state any more, which means inode's run_time flags has _no_ BTRFS_INODE_HAS_ORPHAN_ITEM. Thus, the BUG_ON was triggered because of a wrong check for the flags. Reviewed-by: David Sterba Signed-off-by: Liu Bo Signed-off-by: Josef Bacik diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 6971fb5..9f07bd1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3754,7 +3754,7 @@ void btrfs_evict_inode(struct inode *inode) btrfs_wait_ordered_range(inode, 0, (u64)-1); if (root->fs_info->log_root_recovering) { - BUG_ON(!test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM, + BUG_ON(test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM, &BTRFS_I(inode)->runtime_flags)); goto no_delete; } -- cgit v0.10.2 From bdb7d303b33c1648514c9f9461d7513a4c05ce48 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 27 Jun 2012 15:10:56 -0400 Subject: Btrfs: fix tree log remove space corner case The tree log stuff can have allocated space that we end up having split across a bitmap and a real extent. The free space code does not deal with this, it assumes that if it finds an extent or bitmap entry that the entire range must fall within the entry it finds. This isn't necessarily the case, so rework the remove function so it can handle this case properly. This fixed two panics the user hit, first in the case where the space was initially in a bitmap and then in an extent entry, and then the reverse case. Thanks, Reported-and-tested-by: Shaun Reich Signed-off-by: Josef Bacik diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 19a0d85..a70c54e 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -1542,29 +1542,26 @@ again: end = bitmap_info->offset + (u64)(BITS_PER_BITMAP * ctl->unit) - 1; /* - * XXX - this can go away after a few releases. - * - * since the only user of btrfs_remove_free_space is the tree logging - * stuff, and the only way to test that is under crash conditions, we - * want to have this debug stuff here just in case somethings not - * working. Search the bitmap for the space we are trying to use to - * make sure its actually there. If its not there then we need to stop - * because something has gone wrong. + * We need to search for bits in this bitmap. We could only cover some + * of the extent in this bitmap thanks to how we add space, so we need + * to search for as much as it as we can and clear that amount, and then + * go searching for the next bit. */ search_start = *offset; - search_bytes = *bytes; + search_bytes = ctl->unit; search_bytes = min(search_bytes, end - search_start + 1); ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes); BUG_ON(ret < 0 || search_start != *offset); - if (*offset > bitmap_info->offset && *offset + *bytes > end) { - bitmap_clear_bits(ctl, bitmap_info, *offset, end - *offset + 1); - *bytes -= end - *offset + 1; - *offset = end + 1; - } else if (*offset >= bitmap_info->offset && *offset + *bytes <= end) { - bitmap_clear_bits(ctl, bitmap_info, *offset, *bytes); - *bytes = 0; - } + /* We may have found more bits than what we need */ + search_bytes = min(search_bytes, *bytes); + + /* Cannot clear past the end of the bitmap */ + search_bytes = min(search_bytes, end - search_start + 1); + + bitmap_clear_bits(ctl, bitmap_info, search_start, search_bytes); + *offset += search_bytes; + *bytes -= search_bytes; if (*bytes) { struct rb_node *next = rb_next(&bitmap_info->offset_index); @@ -1595,7 +1592,7 @@ again: * everything over again. */ search_start = *offset; - search_bytes = *bytes; + search_bytes = ctl->unit; ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes); if (ret < 0 || search_start != *offset) @@ -1878,12 +1875,14 @@ int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, { struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *info; - struct btrfs_free_space *next_info = NULL; int ret = 0; spin_lock(&ctl->tree_lock); again: + if (!bytes) + goto out_lock; + info = tree_search_offset(ctl, offset, 0, 0); if (!info) { /* @@ -1904,88 +1903,48 @@ again: } } - if (info->bytes < bytes && rb_next(&info->offset_index)) { - u64 end; - next_info = rb_entry(rb_next(&info->offset_index), - struct btrfs_free_space, - offset_index); - - if (next_info->bitmap) - end = next_info->offset + - BITS_PER_BITMAP * ctl->unit - 1; - else - end = next_info->offset + next_info->bytes; - - if (next_info->bytes < bytes || - next_info->offset > offset || offset > end) { - printk(KERN_CRIT "Found free space at %llu, size %llu," - " trying to use %llu\n", - (unsigned long long)info->offset, - (unsigned long long)info->bytes, - (unsigned long long)bytes); - WARN_ON(1); - ret = -EINVAL; - goto out_lock; - } - - info = next_info; - } - - if (info->bytes == bytes) { + if (!info->bitmap) { unlink_free_space(ctl, info); - if (info->bitmap) { - kfree(info->bitmap); - ctl->total_bitmaps--; - } - kmem_cache_free(btrfs_free_space_cachep, info); - ret = 0; - goto out_lock; - } - - if (!info->bitmap && info->offset == offset) { - unlink_free_space(ctl, info); - info->offset += bytes; - info->bytes -= bytes; - ret = link_free_space(ctl, info); - WARN_ON(ret); - goto out_lock; - } + if (offset == info->offset) { + u64 to_free = min(bytes, info->bytes); + + info->bytes -= to_free; + info->offset += to_free; + if (info->bytes) { + ret = link_free_space(ctl, info); + WARN_ON(ret); + } else { + kmem_cache_free(btrfs_free_space_cachep, info); + } - if (!info->bitmap && info->offset <= offset && - info->offset + info->bytes >= offset + bytes) { - u64 old_start = info->offset; - /* - * we're freeing space in the middle of the info, - * this can happen during tree log replay - * - * first unlink the old info and then - * insert it again after the hole we're creating - */ - unlink_free_space(ctl, info); - if (offset + bytes < info->offset + info->bytes) { - u64 old_end = info->offset + info->bytes; + offset += to_free; + bytes -= to_free; + goto again; + } else { + u64 old_end = info->bytes + info->offset; - info->offset = offset + bytes; - info->bytes = old_end - info->offset; + info->bytes = offset - info->offset; ret = link_free_space(ctl, info); WARN_ON(ret); if (ret) goto out_lock; - } else { - /* the hole we're creating ends at the end - * of the info struct, just free the info - */ - kmem_cache_free(btrfs_free_space_cachep, info); - } - spin_unlock(&ctl->tree_lock); - /* step two, insert a new info struct to cover - * anything before the hole - */ - ret = btrfs_add_free_space(block_group, old_start, - offset - old_start); - WARN_ON(ret); /* -ENOMEM */ - goto out; + /* Not enough bytes in this entry to satisfy us */ + if (old_end < offset + bytes) { + bytes -= old_end - offset; + offset = old_end; + goto again; + } else if (old_end == offset + bytes) { + /* all done */ + goto out_lock; + } + spin_unlock(&ctl->tree_lock); + + ret = btrfs_add_free_space(block_group, offset + bytes, + old_end - (offset + bytes)); + WARN_ON(ret); + goto out; + } } ret = remove_from_bitmap(ctl, info, &offset, &bytes); -- cgit v0.10.2 From 7fd1a3f73f3743b4ffd520effe288a70b0ec47c9 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 27 Jun 2012 17:18:41 -0400 Subject: Btrfs: hold a ref on the inode during writepages We can race with unlink and not actually be able to do our igrab in btrfs_add_ordered_extent. This will result in all sorts of problems. Instead of doing the complicated work to try and handle returning an error properly from btrfs_add_ordered_extent, just hold a ref to the inode during writepages. If we cannot grab a ref we know we're freeing this inode anyway and can just drop the dirty pages on the floor, because screw them we're going to invalidate them anyway. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index aaa12c1..01c21b6 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3324,6 +3324,7 @@ static int extent_write_cache_pages(struct extent_io_tree *tree, writepage_t writepage, void *data, void (*flush_fn)(void *)) { + struct inode *inode = mapping->host; int ret = 0; int done = 0; int nr_to_write_done = 0; @@ -3334,6 +3335,18 @@ static int extent_write_cache_pages(struct extent_io_tree *tree, int scanned = 0; int tag; + /* + * We have to hold onto the inode so that ordered extents can do their + * work when the IO finishes. The alternative to this is failing to add + * an ordered extent if the igrab() fails there and that is a huge pain + * to deal with, so instead just hold onto the inode throughout the + * writepages operation. If it fails here we are freeing up the inode + * anyway and we'd rather not waste our time writing out stuff that is + * going to be truncated anyway. + */ + if (!igrab(inode)) + return 0; + pagevec_init(&pvec, 0); if (wbc->range_cyclic) { index = mapping->writeback_index; /* Start from prev offset */ @@ -3428,6 +3441,7 @@ retry: index = 0; goto retry; } + btrfs_add_delayed_iput(inode); return ret; } -- cgit v0.10.2 From b6305567e7d31b0bec1b8cb9ec0cadd7f7086f5f Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Mon, 2 Jul 2012 15:29:53 -0400 Subject: Btrfs: run delayed directory updates during log replay While we are resolving directory modifications in the tree log, we are triggering delayed metadata updates to the filesystem btrees. This commit forces the delayed updates to run so the replay code can find any modifications done. It stops us from crashing because the directory deleltion replay expects items to be removed immediately from the tree. Signed-off-by: Chris Mason cc: stable@kernel.org diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 2017d0f..8abeae4 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -690,6 +690,8 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans, kfree(name); iput(inode); + + btrfs_run_delayed_items(trans, root); return ret; } @@ -895,6 +897,7 @@ again: ret = btrfs_unlink_inode(trans, root, dir, inode, victim_name, victim_name_len); + btrfs_run_delayed_items(trans, root); } kfree(victim_name); ptr = (unsigned long)(victim_ref + 1) + victim_name_len; @@ -1475,6 +1478,9 @@ again: ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len); BUG_ON(ret); + + btrfs_run_delayed_items(trans, root); + kfree(name); iput(inode); -- cgit v0.10.2 From 8bea2bd37df08aaa599aa361a9f8b836ba98e554 Mon Sep 17 00:00:00 2001 From: Stanislaw Ledwon Date: Mon, 18 Jun 2012 15:20:00 +0200 Subject: usb: Add support for root hub port status CAS The host controller port status register supports CAS (Cold Attach Status) bit. This bit could be set when USB3.0 device is connected when system is in Sx state. When the system wakes to S0 this port status with CAS bit is reported and this port can't be used by any device. When CAS bit is set the port should be reset by warm reset. This was not supported by xhci driver. The issue was found when pendrive was connected to suspended platform. The link state of "Compliance Mode" was reported together with CAS bit. This link state was also not supported by xhci and core/hub.c. The CAS bit is defined only for xhci root hub port and it is not supported on regular hubs. The link status is used to force warm reset on port. Make the USB core issue a warm reset when port is in ether the 'inactive' or 'compliance mode'. Change the xHCI driver to report 'compliance mode' when the CAS is set. This force warm reset on the root hub port. This patch should be backported to stable kernels as old as 3.2, that contain the commit 10d674a82e553cb8a1f41027bb3c3e309b3f6804 "USB: When hot reset for USB3 fails, try warm reset." Signed-off-by: Stanislaw Ledwon Signed-off-by: Sarah Sharp Acked-by: Andiry Xu Cc: stable@vger.kernel.org diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index 25a7422..8fb4849 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -2324,12 +2324,16 @@ static unsigned hub_is_wusb(struct usb_hub *hub) static int hub_port_reset(struct usb_hub *hub, int port1, struct usb_device *udev, unsigned int delay, bool warm); -/* Is a USB 3.0 port in the Inactive state? */ -static bool hub_port_inactive(struct usb_hub *hub, u16 portstatus) +/* Is a USB 3.0 port in the Inactive or Complinance Mode state? + * Port worm reset is required to recover + */ +static bool hub_port_warm_reset_required(struct usb_hub *hub, u16 portstatus) { return hub_is_superspeed(hub->hdev) && - (portstatus & USB_PORT_STAT_LINK_STATE) == - USB_SS_PORT_LS_SS_INACTIVE; + (((portstatus & USB_PORT_STAT_LINK_STATE) == + USB_SS_PORT_LS_SS_INACTIVE) || + ((portstatus & USB_PORT_STAT_LINK_STATE) == + USB_SS_PORT_LS_COMP_MOD)) ; } static int hub_port_wait_reset(struct usb_hub *hub, int port1, @@ -2365,7 +2369,7 @@ static int hub_port_wait_reset(struct usb_hub *hub, int port1, * * See https://bugzilla.kernel.org/show_bug.cgi?id=41752 */ - if (hub_port_inactive(hub, portstatus)) { + if (hub_port_warm_reset_required(hub, portstatus)) { int ret; if ((portchange & USB_PORT_STAT_C_CONNECTION)) @@ -4408,9 +4412,7 @@ static void hub_events(void) /* Warm reset a USB3 protocol port if it's in * SS.Inactive state. */ - if (hub_is_superspeed(hub->hdev) && - (portstatus & USB_PORT_STAT_LINK_STATE) - == USB_SS_PORT_LS_SS_INACTIVE) { + if (hub_port_warm_reset_required(hub, portstatus)) { dev_dbg(hub_dev, "warm reset port %d\n", i); hub_port_reset(hub, i, NULL, HUB_BH_RESET_TIME, true); diff --git a/drivers/usb/host/xhci-hub.c b/drivers/usb/host/xhci-hub.c index 2732ef6..7b01094 100644 --- a/drivers/usb/host/xhci-hub.c +++ b/drivers/usb/host/xhci-hub.c @@ -462,6 +462,42 @@ void xhci_test_and_clear_bit(struct xhci_hcd *xhci, __le32 __iomem **port_array, } } +/* Updates Link Status for super Speed port */ +static void xhci_hub_report_link_state(u32 *status, u32 status_reg) +{ + u32 pls = status_reg & PORT_PLS_MASK; + + /* resume state is a xHCI internal state. + * Do not report it to usb core. + */ + if (pls == XDEV_RESUME) + return; + + /* When the CAS bit is set then warm reset + * should be performed on port + */ + if (status_reg & PORT_CAS) { + /* The CAS bit can be set while the port is + * in any link state. + * Only roothubs have CAS bit, so we + * pretend to be in compliance mode + * unless we're already in compliance + * or the inactive state. + */ + if (pls != USB_SS_PORT_LS_COMP_MOD && + pls != USB_SS_PORT_LS_SS_INACTIVE) { + pls = USB_SS_PORT_LS_COMP_MOD; + } + /* Return also connection bit - + * hub state machine resets port + * when this bit is set. + */ + pls |= USB_PORT_STAT_CONNECTION; + } + /* update status field */ + *status |= pls; +} + int xhci_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue, u16 wIndex, char *buf, u16 wLength) { @@ -606,13 +642,9 @@ int xhci_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue, else status |= USB_PORT_STAT_POWER; } - /* Port Link State */ + /* Update Port Link State for super speed ports*/ if (hcd->speed == HCD_USB3) { - /* resume state is a xHCI internal state. - * Do not report it to usb core. - */ - if ((temp & PORT_PLS_MASK) != XDEV_RESUME) - status |= (temp & PORT_PLS_MASK); + xhci_hub_report_link_state(&status, temp); } if (bus_state->port_c_suspend & (1 << wIndex)) status |= 1 << USB_PORT_FEAT_C_SUSPEND; diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h index de3d6e3..55c0785 100644 --- a/drivers/usb/host/xhci.h +++ b/drivers/usb/host/xhci.h @@ -341,7 +341,11 @@ struct xhci_op_regs { #define PORT_PLC (1 << 22) /* port configure error change - port failed to configure its link partner */ #define PORT_CEC (1 << 23) -/* bit 24 reserved */ +/* Cold Attach Status - xHC can set this bit to report device attached during + * Sx state. Warm port reset should be perfomed to clear this bit and move port + * to connected state. + */ +#define PORT_CAS (1 << 24) /* wake on connect (enable) */ #define PORT_WKCONN_E (1 << 25) /* wake on disconnect (enable) */ -- cgit v0.10.2 From 0d9f78a92ef5e97d9fe51d9215ebe22f6f0d289d Mon Sep 17 00:00:00 2001 From: Sarah Sharp Date: Thu, 21 Jun 2012 16:28:30 -0700 Subject: xhci: Fix hang on back-to-back Set TR Deq Ptr commands. The Microsoft LifeChat 3000 USB headset was causing a very reproducible hang whenever it was plugged in. At first, I thought the host controller was producing bad transfer events, because the log was filled with errors like: xhci_hcd 0000:00:14.0: ERROR Transfer event TRB DMA ptr not part of current TD However, it turned out to be an xHCI driver bug in the ring expansion patches. The bug is triggered When there are two ring segments, and a TD that ends just before a link TRB, like so: ______________ _____________ | | ---> | setup TRB B | ______________ | _____________ | | | | data TRB B | ______________ | _____________ | setup TRB A | <-- deq | | data TRB B | ______________ | _____________ | data TRB A | | | | <-- enq, deq'' ______________ | _____________ | status TRB A | | | | ______________ | _____________ | link TRB |--------------- | link TRB | _____________ <--- deq' _____________ TD A (the first control transfer) stalls on the data phase. That halts the ring. The xHCI driver moves the hardware dequeue pointer to the first TRB after the stalled transfer, which happens to be the link TRB. Once the Set TR dequeue pointer command completes, the function update_ring_for_set_deq_completion runs. That function is supposed to update the xHCI driver's dequeue pointer to match the internal hardware dequeue pointer. On the first call this would work fine, and the software dequeue pointer would move to deq'. However, if the transfer immediately after that stalled (TD B in this case), another Set TR Dequeue command would be issued. That would move the hardware dequeue pointer to deq''. Once that command completed, update_ring_for_set_deq_completion would run again. The original code would unconditionally increment the software dequeue pointer, which moved the pointer off the ring segment into la-la-land. The while loop would happy increment the dequeue pointer (possibly wrapping it) until it matched the hardware pointer value. The while loop would also access all the memory in between the first ring segment and the second ring segment to determine if it was a link TRB. This could cause general protection faults, although it was unlikely because the ring segments came from a DMA pool, and would often have consecutive memory addresses. If nothing in that space looked like a link TRB, the deq_seg pointer for the ring would remain on the first segment. Thus, the deq_seg and the software dequeue pointer would get out of sync. When the next transfer event came in after the stalled transfer, the xHCI driver code would attempt to convert the software dequeue pointer into a DMA address in order to compare the DMA address for the completed transfer. Since the deq_seg and the dequeue pointer were out of sync, xhci_trb_virt_to_dma would return NULL. The transfer event would get ignored, the transfer would eventually timeout, and we would mistakenly convert the finished transfer to no-op TRBs. Some kernel driver (maybe xHCI?) would then get stuck in an infinite loop in interrupt context, and the whole machine would hang. This patch should be backported to kernels as old as 3.4, that contain the commit b008df60c6369ba0290fa7daa177375407a12e07 "xHCI: count free TRBs on transfer ring" Signed-off-by: Sarah Sharp Cc: Andiry Xu Cc: stable@vger.kernel.org diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index 23b4aef..8275645 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -885,6 +885,17 @@ static void update_ring_for_set_deq_completion(struct xhci_hcd *xhci, num_trbs_free_temp = ep_ring->num_trbs_free; dequeue_temp = ep_ring->dequeue; + /* If we get two back-to-back stalls, and the first stalled transfer + * ends just before a link TRB, the dequeue pointer will be left on + * the link TRB by the code in the while loop. So we have to update + * the dequeue pointer one segment further, or we'll jump off + * the segment into la-la-land. + */ + if (last_trb(xhci, ep_ring, ep_ring->deq_seg, ep_ring->dequeue)) { + ep_ring->deq_seg = ep_ring->deq_seg->next; + ep_ring->dequeue = ep_ring->deq_seg->trbs; + } + while (ep_ring->dequeue != dev->eps[ep_index].queued_deq_ptr) { /* We have more usable TRBs */ ep_ring->num_trbs_free++; -- cgit v0.10.2 From d4db2935e4fffeba42540b0dc9d85e3036701221 Mon Sep 17 00:00:00 2001 From: Alex Williamson Date: Fri, 29 Jun 2012 09:56:08 -0600 Subject: KVM: Pass kvm_irqfd to functions Prune this down to just the struct kvm_irqfd so we can avoid changing function definition for every flag or field we use. Signed-off-by: Alex Williamson Acked-by: Cornelia Huck Signed-off-by: Marcelo Tosatti diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index c446435..96c158a 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -815,7 +815,7 @@ static inline void kvm_free_irq_routing(struct kvm *kvm) {} #ifdef CONFIG_HAVE_KVM_EVENTFD void kvm_eventfd_init(struct kvm *kvm); -int kvm_irqfd(struct kvm *kvm, int fd, int gsi, int flags); +int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args); void kvm_irqfd_release(struct kvm *kvm); void kvm_irq_routing_update(struct kvm *, struct kvm_irq_routing_table *); int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args); @@ -824,7 +824,7 @@ int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args); static inline void kvm_eventfd_init(struct kvm *kvm) {} -static inline int kvm_irqfd(struct kvm *kvm, int fd, int gsi, int flags) +static inline int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args) { return -EINVAL; } diff --git a/virt/kvm/eventfd.c b/virt/kvm/eventfd.c index f59c1e8..c307c24 100644 --- a/virt/kvm/eventfd.c +++ b/virt/kvm/eventfd.c @@ -198,7 +198,7 @@ static void irqfd_update(struct kvm *kvm, struct _irqfd *irqfd, } static int -kvm_irqfd_assign(struct kvm *kvm, int fd, int gsi) +kvm_irqfd_assign(struct kvm *kvm, struct kvm_irqfd *args) { struct kvm_irq_routing_table *irq_rt; struct _irqfd *irqfd, *tmp; @@ -212,12 +212,12 @@ kvm_irqfd_assign(struct kvm *kvm, int fd, int gsi) return -ENOMEM; irqfd->kvm = kvm; - irqfd->gsi = gsi; + irqfd->gsi = args->gsi; INIT_LIST_HEAD(&irqfd->list); INIT_WORK(&irqfd->inject, irqfd_inject); INIT_WORK(&irqfd->shutdown, irqfd_shutdown); - file = eventfd_fget(fd); + file = eventfd_fget(args->fd); if (IS_ERR(file)) { ret = PTR_ERR(file); goto fail; @@ -298,19 +298,19 @@ kvm_eventfd_init(struct kvm *kvm) * shutdown any irqfd's that match fd+gsi */ static int -kvm_irqfd_deassign(struct kvm *kvm, int fd, int gsi) +kvm_irqfd_deassign(struct kvm *kvm, struct kvm_irqfd *args) { struct _irqfd *irqfd, *tmp; struct eventfd_ctx *eventfd; - eventfd = eventfd_ctx_fdget(fd); + eventfd = eventfd_ctx_fdget(args->fd); if (IS_ERR(eventfd)) return PTR_ERR(eventfd); spin_lock_irq(&kvm->irqfds.lock); list_for_each_entry_safe(irqfd, tmp, &kvm->irqfds.items, list) { - if (irqfd->eventfd == eventfd && irqfd->gsi == gsi) { + if (irqfd->eventfd == eventfd && irqfd->gsi == args->gsi) { /* * This rcu_assign_pointer is needed for when * another thread calls kvm_irq_routing_update before @@ -338,12 +338,12 @@ kvm_irqfd_deassign(struct kvm *kvm, int fd, int gsi) } int -kvm_irqfd(struct kvm *kvm, int fd, int gsi, int flags) +kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args) { - if (flags & KVM_IRQFD_FLAG_DEASSIGN) - return kvm_irqfd_deassign(kvm, fd, gsi); + if (args->flags & KVM_IRQFD_FLAG_DEASSIGN) + return kvm_irqfd_deassign(kvm, args); - return kvm_irqfd_assign(kvm, fd, gsi); + return kvm_irqfd_assign(kvm, args); } /* diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 7e14068..e98a5ca 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -2047,7 +2047,7 @@ static long kvm_vm_ioctl(struct file *filp, r = -EFAULT; if (copy_from_user(&data, argp, sizeof data)) goto out; - r = kvm_irqfd(kvm, data.fd, data.gsi, data.flags); + r = kvm_irqfd(kvm, &data); break; } case KVM_IOEVENTFD: { -- cgit v0.10.2 From f36992e31284131a978d07bb14e582ce67b46b64 Mon Sep 17 00:00:00 2001 From: Alex Williamson Date: Fri, 29 Jun 2012 09:56:16 -0600 Subject: KVM: Add missing KVM_IRQFD API documentation Signed-off-by: Alex Williamson Acked-by: Michael S. Tsirkin Signed-off-by: Marcelo Tosatti diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt index 9301266..2c99483 100644 --- a/Documentation/virtual/kvm/api.txt +++ b/Documentation/virtual/kvm/api.txt @@ -1930,6 +1930,23 @@ The "pte_enc" field provides a value that can OR'ed into the hash PTE's RPN field (ie, it needs to be shifted left by 12 to OR it into the hash PTE second double word). +4.75 KVM_IRQFD + +Capability: KVM_CAP_IRQFD +Architectures: x86 +Type: vm ioctl +Parameters: struct kvm_irqfd (in) +Returns: 0 on success, -1 on error + +Allows setting an eventfd to directly trigger a guest interrupt. +kvm_irqfd.fd specifies the file descriptor to use as the eventfd and +kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When +an event is tiggered on the eventfd, an interrupt is injected into +the guest using the specified gsi pin. The irqfd is removed using +the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd +and kvm_irqfd.gsi. + + 5. The kvm_run structure ------------------------ -- cgit v0.10.2 From 326cf0334bacfdf43039cf75997a022b9e51535d Mon Sep 17 00:00:00 2001 From: Alex Williamson Date: Fri, 29 Jun 2012 09:56:24 -0600 Subject: KVM: Sanitize KVM_IRQFD flags We only know of one so far. Signed-off-by: Alex Williamson Signed-off-by: Marcelo Tosatti diff --git a/virt/kvm/eventfd.c b/virt/kvm/eventfd.c index c307c24..7d7e2aa 100644 --- a/virt/kvm/eventfd.c +++ b/virt/kvm/eventfd.c @@ -340,6 +340,9 @@ kvm_irqfd_deassign(struct kvm *kvm, struct kvm_irqfd *args) int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args) { + if (args->flags & ~KVM_IRQFD_FLAG_DEASSIGN) + return -EINVAL; + if (args->flags & KVM_IRQFD_FLAG_DEASSIGN) return kvm_irqfd_deassign(kvm, args); -- cgit v0.10.2 From 7b86cef34afceeecf57d4d07f3cfab06f8b60b13 Mon Sep 17 00:00:00 2001 From: Jon Hunter Date: Tue, 3 Jul 2012 11:05:50 -0500 Subject: gpio/omap: fix invalid context restore of gpio bank-0 Currently the gpio _runtime_resume/suspend functions are calling the get_context_loss_count() platform function if the function is populated for a gpio bank. This function is used to determine if the gpio bank logic state needs to be restored due to a power transition. This function will be populated for all banks, but it should only be called for banks that have the "loses_context" variable set. It is pointless to call this if loses_context is false as we know the context will never be lost and will not need restoring. For all OMAP2+ devices gpio bank-0 is in an always-on power domain and so will never lose context. We found that the get_context_loss_count() was being called for bank-0 during the probe and returning 1 instead of 0 indicating that the context had been lost. This was causing the context restore function to be called at probe time for this bank and because the context had never been saved, was restoring an invalid state. This ultimately resulted in a crash [1]. This issue is a regression that was exposed by commit 1b1287032 (gpio/omap: fix missing check in *_runtime_suspend()). There are multiple bugs here that need to be addressed ... 1. Why the always-on power domain returns a context loss count of 1? This needs to be fixed in the power domain code [2]. However, the gpio driver should not assume the loss count is 0 to begin with. 2. The omap gpio driver should never be calling get_context_loss_count for a gpio bank in a always-on domain. This is pointless and adds unneccessary overhead. 3. The OMAP gpio driver assumes that the initial power domain context loss count will be 0 at the time the gpio driver is probed. However, it could be possible that this is not the case and an invalid context restore could be performed during the probe. To avoid this only populate the get_context_loss_count() function pointer after the initial call to pm_runtime_get() has occurred. This will ensure that the first pm_runtime_put() initialised the loss count correctly. This patch addresses issues 2 and 3 above. [1] http://marc.info/?l=linux-omap&m=134065775323775&w=2 [2] http://marc.info/?l=linux-omap&m=134100413303810&w=2 Cc: Kevin Hilman Cc: Grant Likely Cc: Linus Walleij Cc: Tarun Kanti DebBarma Cc: Franky Lin Cc: Santosh Shilimkar Cc: NeilBrown Reported-by: Franky Lin Reviewed-by: Santosh Shilimkar Tested-by: Franky Lin Acked-by: Kevin Hilman Tested-by: NeilBrown Signed-off-by: Jon Hunter Signed-off-by: Kevin Hilman diff --git a/drivers/gpio/gpio-omap.c b/drivers/gpio/gpio-omap.c index ff213e7..4fbc208 100644 --- a/drivers/gpio/gpio-omap.c +++ b/drivers/gpio/gpio-omap.c @@ -1091,7 +1091,6 @@ static int __devinit omap_gpio_probe(struct platform_device *pdev) bank->is_mpuio = pdata->is_mpuio; bank->non_wakeup_gpios = pdata->non_wakeup_gpios; bank->loses_context = pdata->loses_context; - bank->get_context_loss_count = pdata->get_context_loss_count; bank->regs = pdata->regs; #ifdef CONFIG_OF_GPIO bank->chip.of_node = of_node_get(node); @@ -1145,6 +1144,9 @@ static int __devinit omap_gpio_probe(struct platform_device *pdev) omap_gpio_chip_init(bank); omap_gpio_show_rev(bank); + if (bank->loses_context) + bank->get_context_loss_count = pdata->get_context_loss_count; + pm_runtime_put(bank->dev); list_add_tail(&bank->node, &omap_gpio_list); -- cgit v0.10.2 From ec01d738a1691dfc85b96b9f796020267a7be577 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 2 Jul 2012 07:24:25 -0400 Subject: cifs: when server doesn't set CAP_LARGE_READ_X, cap default rsize at MaxBufferSize When the server doesn't advertise CAP_LARGE_READ_X, then MS-CIFS states that you must cap the size of the read at the client's MaxBufferSize. Unfortunately, testing with many older servers shows that they often can't service a read larger than their own MaxBufferSize. Since we can't assume what the server will do in this situation, we must be conservative here for the default. When the server can't do large reads, then assume that it can't satisfy any read larger than its MaxBufferSize either. Luckily almost all modern servers can do large reads, so this won't affect them. This is really just for older win9x and OS/2 era servers. Also, note that this patch just governs the default rsize. The admin can always override this if he so chooses. Cc: # 3.2 Reported-by: David H. Durgee Signed-off-by: Jeff Layton Signed-off-by: Steven French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 5b38407..0ae86dd 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -3495,18 +3495,15 @@ cifs_negotiate_rsize(struct cifs_tcon *tcon, struct smb_vol *pvolume_info) * MS-CIFS indicates that servers are only limited by the client's * bufsize for reads, testing against win98se shows that it throws * INVALID_PARAMETER errors if you try to request too large a read. + * OS/2 just sends back short reads. * - * If the server advertises a MaxBufferSize of less than one page, - * assume that it also can't satisfy reads larger than that either. - * - * FIXME: Is there a better heuristic for this? + * If the server doesn't advertise CAP_LARGE_READ_X, then assume that + * it can't handle a read request larger than its MaxBufferSize either. */ if (tcon->unix_ext && (unix_cap & CIFS_UNIX_LARGE_READ_CAP)) defsize = CIFS_DEFAULT_IOSIZE; else if (server->capabilities & CAP_LARGE_READ_X) defsize = CIFS_DEFAULT_NON_POSIX_RSIZE; - else if (server->maxBuf >= PAGE_CACHE_SIZE) - defsize = CIFSMaxBufSize; else defsize = server->maxBuf - sizeof(READ_RSP); -- cgit v0.10.2 From d92d95b6bf2722ffa0fefa7651c51bf336743dd7 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Mon, 2 Jul 2012 19:21:06 -0700 Subject: regulator: Fix recursive mutex lockdep warning A recursive lockdep warning occurs if you call regulator_set_optimum_mode() on a regulator with a supply because there is no nesting annotation for the rdev->mutex. To avoid this warning, get the supply's load before locking the regulator's mutex to avoid grabbing the same class of lock twice. ============================================= [ INFO: possible recursive locking detected ] 3.4.0 #3257 Tainted: G W --------------------------------------------- swapper/0/1 is trying to acquire lock: (&rdev->mutex){+.+.+.}, at: [] regulator_get_voltage+0x18/0x38 but task is already holding lock: (&rdev->mutex){+.+.+.}, at: [] regulator_set_optimum_mode+0x24/0x224 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&rdev->mutex); lock(&rdev->mutex); *** DEADLOCK *** May be due to missing lock nesting notation 3 locks held by swapper/0/1: #0: (&__lockdep_no_validate__){......}, at: [] __driver_attach+0x40/0x8c #1: (&__lockdep_no_validate__){......}, at: [] __driver_attach+0x50/0x8c #2: (&rdev->mutex){+.+.+.}, at: [] regulator_set_optimum_mode+0x24/0x224 stack backtrace: [] (unwind_backtrace+0x0/0x12c) from [] (validate_chain+0x760/0x1080) [] (validate_chain+0x760/0x1080) from [] (__lock_acquire+0x950/0xa10) [] (__lock_acquire+0x950/0xa10) from [] (lock_acquire+0x18c/0x1e8) [] (lock_acquire+0x18c/0x1e8) from [] (mutex_lock_nested+0x68/0x3c4) [] (mutex_lock_nested+0x68/0x3c4) from [] (regulator_get_voltage+0x18/0x38) [] (regulator_get_voltage+0x18/0x38) from [] (regulator_set_optimum_mode+0xa4/0x224) ... Signed-off-by: Stephen Boyd Signed-off-by: Mark Brown diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c index 09a737c..8b4b382 100644 --- a/drivers/regulator/core.c +++ b/drivers/regulator/core.c @@ -2519,9 +2519,12 @@ int regulator_set_optimum_mode(struct regulator *regulator, int uA_load) { struct regulator_dev *rdev = regulator->rdev; struct regulator *consumer; - int ret, output_uV, input_uV, total_uA_load = 0; + int ret, output_uV, input_uV = 0, total_uA_load = 0; unsigned int mode; + if (rdev->supply) + input_uV = regulator_get_voltage(rdev->supply); + mutex_lock(&rdev->mutex); /* @@ -2554,10 +2557,7 @@ int regulator_set_optimum_mode(struct regulator *regulator, int uA_load) goto out; } - /* get input voltage */ - input_uV = 0; - if (rdev->supply) - input_uV = regulator_get_voltage(rdev->supply); + /* No supply? Use constraint voltage */ if (input_uV <= 0) input_uV = rdev->constraints->input_uV; if (input_uV <= 0) { -- cgit v0.10.2 From f411930442e01f9cf1bf4df41ff7e89476575c4d Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Tue, 3 Jul 2012 14:31:01 +0800 Subject: KVM: fix fault page leak fault_page is forgot to be freed Signed-off-by: Xiao Guangrong Signed-off-by: Marcelo Tosatti diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index e98a5ca..44ee712 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -2845,6 +2845,7 @@ void kvm_exit(void) kvm_arch_hardware_unsetup(); kvm_arch_exit(); free_cpumask_var(cpus_hardware_enabled); + __free_page(fault_page); __free_page(hwpoison_page); __free_page(bad_page); } -- cgit v0.10.2 From 85b7059169e128c57a3a8a3e588fb89cb2031da1 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Tue, 3 Jul 2012 14:31:43 +0800 Subject: KVM: MMU: fix shrinking page from the empty mmu Fix: [ 3190.059226] BUG: unable to handle kernel NULL pointer dereference at (null) [ 3190.062224] IP: [] mmu_page_zap_pte+0x10/0xa7 [kvm] [ 3190.063760] PGD 104f50067 PUD 112bea067 PMD 0 [ 3190.065309] Oops: 0000 [#1] SMP DEBUG_PAGEALLOC [ 3190.066860] CPU 1 [ ...... ] [ 3190.109629] Call Trace: [ 3190.111342] [] kvm_mmu_prepare_zap_page+0xa9/0x1fc [kvm] [ 3190.113091] [] mmu_shrink+0x11f/0x1f3 [kvm] [ 3190.114844] [] ? mmu_shrink+0x87/0x1f3 [kvm] [ 3190.116598] [] ? prune_super+0x142/0x154 [ 3190.118333] [] ? shrink_slab+0x39/0x31e [ 3190.120043] [] shrink_slab+0x1cc/0x31e [ 3190.121718] [] do_try_to_free_pages This is caused by shrinking page from the empty mmu, although we have checked n_used_mmu_pages, it is useless since the check is out of mmu-lock Signed-off-by: Xiao Guangrong Signed-off-by: Marcelo Tosatti diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index be3cea4..57e168e 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -3934,6 +3934,9 @@ static void kvm_mmu_remove_some_alloc_mmu_pages(struct kvm *kvm, { struct kvm_mmu_page *page; + if (list_empty(&kvm->arch.active_mmu_pages)) + return; + page = container_of(kvm->arch.active_mmu_pages.prev, struct kvm_mmu_page, link); kvm_mmu_prepare_zap_page(kvm, page, invalid_list); -- cgit v0.10.2 From 9fe79d7600497ed8a95c3981cbe5b73ab98222f0 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Tue, 12 Jun 2012 11:17:01 -0700 Subject: eCryptfs: Properly check for O_RDONLY flag before doing privileged open If the first attempt at opening the lower file read/write fails, eCryptfs will retry using a privileged kthread. However, the privileged retry should not happen if the lower file's inode is read-only because a read/write open will still be unsuccessful. The check for determining if the open should be retried was intended to be based on the access mode of the lower file's open flags being O_RDONLY, but the check was incorrectly performed. This would cause the open to be retried by the privileged kthread, resulting in a second failed open of the lower file. This patch corrects the check to determine if the open request should be handled by the privileged kthread. Signed-off-by: Tyler Hicks Reported-by: Dan Carpenter Acked-by: Dan Carpenter diff --git a/fs/ecryptfs/kthread.c b/fs/ecryptfs/kthread.c index 69f994a..0dbe58a 100644 --- a/fs/ecryptfs/kthread.c +++ b/fs/ecryptfs/kthread.c @@ -149,7 +149,7 @@ int ecryptfs_privileged_open(struct file **lower_file, (*lower_file) = dentry_open(lower_dentry, lower_mnt, flags, cred); if (!IS_ERR(*lower_file)) goto out; - if (flags & O_RDONLY) { + if ((flags & O_ACCMODE) == O_RDONLY) { rc = PTR_ERR((*lower_file)); goto out; } -- cgit v0.10.2 From 60d65f1f07a7d81d3eb3b91fc13fca80f2fdbb12 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Mon, 11 Jun 2012 10:21:34 -0700 Subject: eCryptfs: Fix lockdep warning in miscdev operations Don't grab the daemon mutex while holding the message context mutex. Addresses this lockdep warning: ecryptfsd/2141 is trying to acquire lock: (&ecryptfs_msg_ctx_arr[i].mux){+.+.+.}, at: [] ecryptfs_miscdev_read+0x143/0x470 [ecryptfs] but task is already holding lock: (&(*daemon)->mux){+.+...}, at: [] ecryptfs_miscdev_read+0x21c/0x470 [ecryptfs] which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (&(*daemon)->mux){+.+...}: [] lock_acquire+0x9d/0x220 [] __mutex_lock_common+0x5a/0x4b0 [] mutex_lock_nested+0x44/0x50 [] ecryptfs_send_miscdev+0x97/0x120 [ecryptfs] [] ecryptfs_send_message+0x134/0x1e0 [ecryptfs] [] ecryptfs_generate_key_packet_set+0x2fe/0xa80 [ecryptfs] [] ecryptfs_write_metadata+0x108/0x250 [ecryptfs] [] ecryptfs_create+0x130/0x250 [ecryptfs] [] vfs_create+0xb4/0x120 [] do_last+0x8c5/0xa10 [] path_openat+0xd9/0x460 [] do_filp_open+0x42/0xa0 [] do_sys_open+0xf8/0x1d0 [] sys_open+0x21/0x30 [] system_call_fastpath+0x16/0x1b -> #0 (&ecryptfs_msg_ctx_arr[i].mux){+.+.+.}: [] __lock_acquire+0x1bf8/0x1c50 [] lock_acquire+0x9d/0x220 [] __mutex_lock_common+0x5a/0x4b0 [] mutex_lock_nested+0x44/0x50 [] ecryptfs_miscdev_read+0x143/0x470 [ecryptfs] [] vfs_read+0xb3/0x180 [] sys_read+0x4d/0x90 [] system_call_fastpath+0x16/0x1b Signed-off-by: Tyler Hicks diff --git a/fs/ecryptfs/miscdev.c b/fs/ecryptfs/miscdev.c index 3a06f40..3c632ec 100644 --- a/fs/ecryptfs/miscdev.c +++ b/fs/ecryptfs/miscdev.c @@ -191,31 +191,32 @@ int ecryptfs_send_miscdev(char *data, size_t data_size, struct ecryptfs_msg_ctx *msg_ctx, u8 msg_type, u16 msg_flags, struct ecryptfs_daemon *daemon) { - int rc = 0; + struct ecryptfs_message *msg; - mutex_lock(&msg_ctx->mux); - msg_ctx->msg = kmalloc((sizeof(*msg_ctx->msg) + data_size), - GFP_KERNEL); - if (!msg_ctx->msg) { - rc = -ENOMEM; + msg = kmalloc((sizeof(*msg) + data_size), GFP_KERNEL); + if (!msg) { printk(KERN_ERR "%s: Out of memory whilst attempting " "to kmalloc(%zd, GFP_KERNEL)\n", __func__, - (sizeof(*msg_ctx->msg) + data_size)); - goto out_unlock; + (sizeof(*msg) + data_size)); + return -ENOMEM; } + + mutex_lock(&msg_ctx->mux); + msg_ctx->msg = msg; msg_ctx->msg->index = msg_ctx->index; msg_ctx->msg->data_len = data_size; msg_ctx->type = msg_type; memcpy(msg_ctx->msg->data, data, data_size); msg_ctx->msg_size = (sizeof(*msg_ctx->msg) + data_size); - mutex_lock(&daemon->mux); list_add_tail(&msg_ctx->daemon_out_list, &daemon->msg_ctx_out_queue); + mutex_unlock(&msg_ctx->mux); + + mutex_lock(&daemon->mux); daemon->num_queued_msg_ctx++; wake_up_interruptible(&daemon->wait); mutex_unlock(&daemon->mux); -out_unlock: - mutex_unlock(&msg_ctx->mux); - return rc; + + return 0; } /* -- cgit v0.10.2 From 3e5d3c35a68c9a933bdbdd8685bd1a205b57e806 Mon Sep 17 00:00:00 2001 From: Junxiao Bi Date: Wed, 27 Jun 2012 17:09:55 +0800 Subject: ocfs2: clear unaligned io flag when dio fails The unaligned io flag is set in the kiocb when an unaligned dio is issued, it should be cleared even when the dio fails, or it may affect the following io which are using the same kiocb. Signed-off-by: Junxiao Bi Cc: stable@vger.kernel.org Signed-off-by: Joel Becker diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 061591a..98513c8 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2422,8 +2422,10 @@ out_dio: unaligned_dio = 0; } - if (unaligned_dio) + if (unaligned_dio) { + ocfs2_iocb_clear_unaligned_aio(iocb); atomic_dec(&OCFS2_I(inode)->ip_unaligned_aio); + } out: if (rw_level != -1) -- cgit v0.10.2 From 16865b7c42fbce8a4d2b278460e387e719e289cb Mon Sep 17 00:00:00 2001 From: roel Date: Mon, 12 Dec 2011 23:40:51 +0100 Subject: ocfs2: Misplaced parens in unlikley Fix misplaced parentheses Signed-off-by: Roel Kluin Signed-off-by: Joel Becker diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 81a4cd2..274529c 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -456,7 +456,7 @@ static void ocfs2_update_lock_stats(struct ocfs2_lock_res *res, int level, stats->ls_gets++; stats->ls_total += ktime_to_ns(kt); /* overflow */ - if (unlikely(stats->ls_gets) == 0) { + if (unlikely(stats->ls_gets == 0)) { stats->ls_gets++; stats->ls_total = ktime_to_ns(kt); } -- cgit v0.10.2 From a75e9ccabd925d16954739bd977c54695c9310d0 Mon Sep 17 00:00:00 2001 From: Srinivas Eeda Date: Mon, 30 Jan 2012 21:51:22 -0800 Subject: ocfs2: use spinlock irqsave for downconvert lock.patch When ocfs2dc thread holds dc_task_lock spinlock and receives soft IRQ it deadlock itself trying to get same spinlock in ocfs2_wake_downconvert_thread. Below is the stack snippet. The patch disables interrupts when acquiring dc_task_lock spinlock. ocfs2_wake_downconvert_thread ocfs2_rw_unlock ocfs2_dio_end_io dio_complete ..... bio_endio req_bio_endio .... scsi_io_completion blk_done_softirq __do_softirq do_softirq irq_exit do_IRQ ocfs2_downconvert_thread [kthread] Signed-off-by: Srinivas Eeda Signed-off-by: Joel Becker diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 274529c..4f7795f 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -3932,6 +3932,8 @@ unqueue: static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb, struct ocfs2_lock_res *lockres) { + unsigned long flags; + assert_spin_locked(&lockres->l_lock); if (lockres->l_flags & OCFS2_LOCK_FREEING) { @@ -3945,21 +3947,22 @@ static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb, lockres_or_flags(lockres, OCFS2_LOCK_QUEUED); - spin_lock(&osb->dc_task_lock); + spin_lock_irqsave(&osb->dc_task_lock, flags); if (list_empty(&lockres->l_blocked_list)) { list_add_tail(&lockres->l_blocked_list, &osb->blocked_lock_list); osb->blocked_lock_count++; } - spin_unlock(&osb->dc_task_lock); + spin_unlock_irqrestore(&osb->dc_task_lock, flags); } static void ocfs2_downconvert_thread_do_work(struct ocfs2_super *osb) { unsigned long processed; + unsigned long flags; struct ocfs2_lock_res *lockres; - spin_lock(&osb->dc_task_lock); + spin_lock_irqsave(&osb->dc_task_lock, flags); /* grab this early so we know to try again if a state change and * wake happens part-way through our work */ osb->dc_work_sequence = osb->dc_wake_sequence; @@ -3972,38 +3975,40 @@ static void ocfs2_downconvert_thread_do_work(struct ocfs2_super *osb) struct ocfs2_lock_res, l_blocked_list); list_del_init(&lockres->l_blocked_list); osb->blocked_lock_count--; - spin_unlock(&osb->dc_task_lock); + spin_unlock_irqrestore(&osb->dc_task_lock, flags); BUG_ON(!processed); processed--; ocfs2_process_blocked_lock(osb, lockres); - spin_lock(&osb->dc_task_lock); + spin_lock_irqsave(&osb->dc_task_lock, flags); } - spin_unlock(&osb->dc_task_lock); + spin_unlock_irqrestore(&osb->dc_task_lock, flags); } static int ocfs2_downconvert_thread_lists_empty(struct ocfs2_super *osb) { int empty = 0; + unsigned long flags; - spin_lock(&osb->dc_task_lock); + spin_lock_irqsave(&osb->dc_task_lock, flags); if (list_empty(&osb->blocked_lock_list)) empty = 1; - spin_unlock(&osb->dc_task_lock); + spin_unlock_irqrestore(&osb->dc_task_lock, flags); return empty; } static int ocfs2_downconvert_thread_should_wake(struct ocfs2_super *osb) { int should_wake = 0; + unsigned long flags; - spin_lock(&osb->dc_task_lock); + spin_lock_irqsave(&osb->dc_task_lock, flags); if (osb->dc_work_sequence != osb->dc_wake_sequence) should_wake = 1; - spin_unlock(&osb->dc_task_lock); + spin_unlock_irqrestore(&osb->dc_task_lock, flags); return should_wake; } @@ -4033,10 +4038,12 @@ static int ocfs2_downconvert_thread(void *arg) void ocfs2_wake_downconvert_thread(struct ocfs2_super *osb) { - spin_lock(&osb->dc_task_lock); + unsigned long flags; + + spin_lock_irqsave(&osb->dc_task_lock, flags); /* make sure the voting thread gets a swipe at whatever changes * the caller may have made to the voting state */ osb->dc_wake_sequence++; - spin_unlock(&osb->dc_task_lock); + spin_unlock_irqrestore(&osb->dc_task_lock, flags); wake_up(&osb->dc_event); } -- cgit v0.10.2 From 65622e647bfff3ed89f95576ec120693ed4085a6 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Thu, 9 Feb 2012 14:42:22 +0800 Subject: ocfs2: for SEEK_DATA/SEEK_HOLE, return internal error unchanged if ocfs2_get_clusters_nocache() or ocfs2_inode_lock() call failed. Hello, Since ENXIO only means "offset beyond EOF" for SEEK_DATA/SEEK_HOLE, Hence we should return the internal error unchanged if ocfs2_inode_lock() or ocfs2_get_clusters_nocache() call failed rather than ENXIO. Otherwise, it will confuse the user applications when they trying to understand the root cause. Thanks Dave for pointing this out. Thanks, -Jeff Cc: Dave Chinner Signed-off-by: Jie Liu Signed-off-by: Joel Becker diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c index 2f5b92e..70b5863 100644 --- a/fs/ocfs2/extent_map.c +++ b/fs/ocfs2/extent_map.c @@ -923,8 +923,6 @@ out_unlock: ocfs2_inode_unlock(inode, 0); out: - if (ret && ret != -ENXIO) - ret = -ENXIO; return ret; } -- cgit v0.10.2 From a4564ead763a9264edbec6d4e72aa273f05eb39c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 10 Feb 2012 10:50:07 +0100 Subject: ocfs2: Fix bogus error message from ocfs2_global_read_info 'status' variable in ocfs2_global_read_info() is always != 0 when leaving the function because it happens to contain number of read bytes. Thus we always log error message although everything is OK. Since all error cases properly call mlog_errno() before jumping to out_err, there's no reason to call mlog_errno() on exit at all. This is a fallout of c1e8d35e (conversion of mlog_exit() calls). Signed-off-by: Jan Kara Signed-off-by: Joel Becker diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index 92fcd57..0a86e30 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -399,8 +399,6 @@ int ocfs2_global_read_info(struct super_block *sb, int type) msecs_to_jiffies(oinfo->dqi_syncms)); out_err: - if (status) - mlog_errno(status); return status; out_unlock: ocfs2_unlock_global_qf(oinfo, 0); -- cgit v0.10.2 From c520c921eacdced7e2095ba6cbbb9921906c7b67 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Mon, 2 Jul 2012 11:30:34 +0200 Subject: ARM: imx: assert SCC gate stays enabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The SCC clock is needed in internal boot mode and so must keep enabled. This same issue was fixed for the pre-common-clk code in commit 3d6e614 (mx35: Fix boot ROM hang in internal boot mode) Cc: John Ogness Cc: Hans J. Koch Signed-off-by: Uwe Kleine-König Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-imx/clk-imx35.c b/arch/arm/mach-imx/clk-imx35.c index 920a8cc..c6422fb 100644 --- a/arch/arm/mach-imx/clk-imx35.c +++ b/arch/arm/mach-imx/clk-imx35.c @@ -201,7 +201,6 @@ int __init mx35_clocks_init() pr_err("i.MX35 clk %d: register failed with %ld\n", i, PTR_ERR(clk[i])); - clk_register_clkdev(clk[pata_gate], NULL, "pata_imx"); clk_register_clkdev(clk[can1_gate], NULL, "flexcan.0"); clk_register_clkdev(clk[can2_gate], NULL, "flexcan.1"); @@ -264,6 +263,14 @@ int __init mx35_clocks_init() clk_prepare_enable(clk[iim_gate]); clk_prepare_enable(clk[emi_gate]); + /* + * SCC is needed to boot via mmc after a watchdog reset. The clock code + * before conversion to common clk also enabled UART1 (which isn't + * handled here and not needed for mmc) and IIM (which is enabled + * unconditionally above). + */ + clk_prepare_enable(clk[scc_gate]); + imx_print_silicon_rev("i.MX35", mx35_revision()); #ifdef CONFIG_MXC_USE_EPIT -- cgit v0.10.2 From d50394266b340d930a7458fa669d36e99670f200 Mon Sep 17 00:00:00 2001 From: Ohad Ben-Cohen Date: Sun, 1 Jul 2012 11:31:35 +0300 Subject: remoteproc/omap: fix randconfig unmet direct dependencies OMAP_REMOTEPROC selects REMOTEPROC and RPMSG, both of which depend on EXPERIMENTAL, so let's have OMAP_REMOTEPROC depend on EXPERIMENTAL too, in order to avoid the below randconfig warnings. warning: (OMAP_REMOTEPROC) selects REMOTEPROC which has unmet direct dependencies (EXPERIMENTAL) warning: (OMAP_REMOTEPROC) selects RPMSG which has unmet direct dependencies (EXPERIMENTAL) Cc: stable Reported-by: Tony Lindgren Signed-off-by: Ohad Ben-Cohen diff --git a/drivers/remoteproc/Kconfig b/drivers/remoteproc/Kconfig index 24d880e..db4e39b 100644 --- a/drivers/remoteproc/Kconfig +++ b/drivers/remoteproc/Kconfig @@ -7,6 +7,7 @@ config REMOTEPROC config OMAP_REMOTEPROC tristate "OMAP remoteproc support" + depends on EXPERIMENTAL depends on ARCH_OMAP4 depends on OMAP_IOMMU select REMOTEPROC -- cgit v0.10.2 From fb31fbeb647ed1910290712c13f11f1db09e5506 Mon Sep 17 00:00:00 2001 From: Alexander Holler Date: Tue, 3 Jul 2012 14:35:47 +0800 Subject: leds: heartbeat: fix bug on panic With commit 49dca5aebfdeadd4bf27b6cb4c60392147dc35a4 I introduced a bug (visible if CONFIG_PROVE_RCU is enabled) which occures when a panic has happened: [ 1526.520230] =============================== [ 1526.520230] [ INFO: suspicious RCU usage. ] [ 1526.520230] 3.5.0-rc1+ #12 Not tainted [ 1526.520230] ------------------------------- [ 1526.520230] /c/kernel-tests/mm/include/linux/rcupdate.h:436 Illegal context switch in RCU read-side critical section! [ 1526.520230] [ 1526.520230] other info that might help us debug this: [ 1526.520230] [ 1526.520230] [ 1526.520230] rcu_scheduler_active = 1, debug_locks = 0 [ 1526.520230] 3 locks held by net.agent/3279: [ 1526.520230] #0: (&mm->mmap_sem){++++++}, at: [] do_page_fault+0x193/0x390 [ 1526.520230] #1: (panic_lock){+.+...}, at: [] panic+0x37/0x1d3 [ 1526.520230] #2: (rcu_read_lock){.+.+..}, at: [] rcu_lock_acquire+0x0/0x29 [ 1526.520230] [ 1526.520230] stack backtrace: [ 1526.520230] Pid: 3279, comm: net.agent Not tainted 3.5.0-rc1+ #12 [ 1526.520230] Call Trace: [ 1526.520230] [] lockdep_rcu_suspicious+0x109/0x112 [ 1526.520230] [] rcu_preempt_sleep_check+0x45/0x47 [ 1526.520230] [] __might_sleep+0x1e/0x19a [ 1526.520230] [] down_write+0x26/0x81 [ 1526.520230] [] led_trigger_unregister+0x1f/0x9c [ 1526.520230] [] heartbeat_reboot_notifier+0x15/0x19 [ 1526.520230] [] notifier_call_chain+0x96/0xcd [ 1526.520230] [] __atomic_notifier_call_chain+0x8e/0xff [ 1526.520230] [] ? kmsg_dump+0x37/0x1eb [ 1526.520230] [] atomic_notifier_call_chain+0x14/0x16 [ 1526.520230] [] panic+0xe8/0x1d3 [ 1526.520230] [] out_of_memory+0x15d/0x1d3 So in case of a panic, now just turn of the LED. Other approaches like scheduling a work to unregister the trigger aren't working because there isn't much which still runs after a panic occured (except timers). Signed-off-by: Alexander Holler Signed-off-by: Bryan Wu diff --git a/drivers/leds/ledtrig-heartbeat.c b/drivers/leds/ledtrig-heartbeat.c index 41dc76d..a019fbb 100644 --- a/drivers/leds/ledtrig-heartbeat.c +++ b/drivers/leds/ledtrig-heartbeat.c @@ -21,6 +21,8 @@ #include #include "leds.h" +static int panic_heartbeats; + struct heartbeat_trig_data { unsigned int phase; unsigned int period; @@ -34,6 +36,11 @@ static void led_heartbeat_function(unsigned long data) unsigned long brightness = LED_OFF; unsigned long delay = 0; + if (unlikely(panic_heartbeats)) { + led_set_brightness(led_cdev, LED_OFF); + return; + } + /* acts like an actual heart beat -- ie thump-thump-pause... */ switch (heartbeat_data->phase) { case 0: @@ -111,12 +118,19 @@ static int heartbeat_reboot_notifier(struct notifier_block *nb, return NOTIFY_DONE; } +static int heartbeat_panic_notifier(struct notifier_block *nb, + unsigned long code, void *unused) +{ + panic_heartbeats = 1; + return NOTIFY_DONE; +} + static struct notifier_block heartbeat_reboot_nb = { .notifier_call = heartbeat_reboot_notifier, }; static struct notifier_block heartbeat_panic_nb = { - .notifier_call = heartbeat_reboot_notifier, + .notifier_call = heartbeat_panic_notifier, }; static int __init heartbeat_trig_init(void) -- cgit v0.10.2 From c12a3cb9dca96768a5f048bb50f3395216346bda Mon Sep 17 00:00:00 2001 From: Paul Bolle Date: Mon, 2 Jul 2012 23:40:14 +0200 Subject: ARM: mmp: remove mach/gpio-pxa.h Commit 157d2644cb0c1e71a18baaffca56d2b1d0ebf10f ("ARM: pxa: change gpio to platform device") removed all includes of mach/gpio-pxa.h. It kept this unused header in the tree. Using it can't work, as it itself includes the non-existent header plat/gpio-pxa.h. This header can safely be removed. Signed-off-by: Paul Bolle Acked-by: Eric Miao Signed-off-by: Haojian Zhuang diff --git a/arch/arm/mach-mmp/include/mach/gpio-pxa.h b/arch/arm/mach-mmp/include/mach/gpio-pxa.h deleted file mode 100644 index 0e135a5..0000000 --- a/arch/arm/mach-mmp/include/mach/gpio-pxa.h +++ /dev/null @@ -1,29 +0,0 @@ -#ifndef __ASM_MACH_GPIO_PXA_H -#define __ASM_MACH_GPIO_PXA_H - -#include -#include -#include - -#define GPIO_REGS_VIRT (APB_VIRT_BASE + 0x19000) - -#define BANK_OFF(n) (((n) < 3) ? (n) << 2 : 0x100 + (((n) - 3) << 2)) -#define GPIO_REG(x) (*(volatile u32 *)(GPIO_REGS_VIRT + (x))) - -#define gpio_to_bank(gpio) ((gpio) >> 5) - -/* NOTE: these macros are defined here to make optimization of - * gpio_{get,set}_value() to work when 'gpio' is a constant. - * Usage of these macros otherwise is no longer recommended, - * use generic GPIO API whenever possible. - */ -#define GPIO_bit(gpio) (1 << ((gpio) & 0x1f)) - -#define GPLR(x) GPIO_REG(BANK_OFF(gpio_to_bank(x)) + 0x00) -#define GPDR(x) GPIO_REG(BANK_OFF(gpio_to_bank(x)) + 0x0c) -#define GPSR(x) GPIO_REG(BANK_OFF(gpio_to_bank(x)) + 0x18) -#define GPCR(x) GPIO_REG(BANK_OFF(gpio_to_bank(x)) + 0x24) - -#include - -#endif /* __ASM_MACH_GPIO_PXA_H */ -- cgit v0.10.2 From e121aefa7d9f10eee5cf26ed47129237a05d940b Mon Sep 17 00:00:00 2001 From: Ohad Ben-Cohen Date: Sun, 1 Jul 2012 11:53:36 +0300 Subject: remoteproc: fix missing CONFIG_FW_LOADER configurations Remoteproc requires user space firmware loading support, so let's select FW_LOADER explicitly to avoid painful misconfigurations (which only show up in runtime). Cc: stable Reported-by: Mark Grosen Signed-off-by: Ohad Ben-Cohen diff --git a/drivers/remoteproc/Kconfig b/drivers/remoteproc/Kconfig index db4e39b..f8d818a 100644 --- a/drivers/remoteproc/Kconfig +++ b/drivers/remoteproc/Kconfig @@ -4,6 +4,7 @@ menu "Remoteproc drivers (EXPERIMENTAL)" config REMOTEPROC tristate depends on EXPERIMENTAL + select FW_CONFIG config OMAP_REMOTEPROC tristate "OMAP remoteproc support" -- cgit v0.10.2 From 1e0c1ce00d83834d03f4d8d039734ca4703298df Mon Sep 17 00:00:00 2001 From: Andrew Lunn Date: Fri, 29 Jun 2012 09:25:58 +0200 Subject: ARM: Orion: Fix WDT compile for Dove and MV78xx0 Commit 0fa1f0609a0c1fe8b2be3c0089a2cb48f7fda521 (ARM: Orion: Fix Virtual/Physical mixup with watchdog) broke the Dove & MV78xx0 build. Although these two SoC don't use the watchdog, the shared platform code still needs to build. Add the necessary defines. Cc: stable@vger.kernel.org Reported-by: Nicolas Pitre Signed-off-by: Andrew Lunn Tested-by: Nicolas Pitre Signed-off-by: Arnd Bergmann diff --git a/arch/arm/mach-dove/include/mach/bridge-regs.h b/arch/arm/mach-dove/include/mach/bridge-regs.h index 226949d..f953bb5 100644 --- a/arch/arm/mach-dove/include/mach/bridge-regs.h +++ b/arch/arm/mach-dove/include/mach/bridge-regs.h @@ -50,5 +50,6 @@ #define POWER_MANAGEMENT (BRIDGE_VIRT_BASE | 0x011c) #define TIMER_VIRT_BASE (BRIDGE_VIRT_BASE | 0x0300) +#define TIMER_PHYS_BASE (BRIDGE_PHYS_BASE | 0x0300) #endif diff --git a/arch/arm/mach-dove/include/mach/dove.h b/arch/arm/mach-dove/include/mach/dove.h index ad1165d..d52b0ef 100644 --- a/arch/arm/mach-dove/include/mach/dove.h +++ b/arch/arm/mach-dove/include/mach/dove.h @@ -78,6 +78,7 @@ /* North-South Bridge */ #define BRIDGE_VIRT_BASE (DOVE_SB_REGS_VIRT_BASE | 0x20000) +#define BRIDGE_PHYS_BASE (DOVE_SB_REGS_PHYS_BASE | 0x20000) /* Cryptographic Engine */ #define DOVE_CRYPT_PHYS_BASE (DOVE_SB_REGS_PHYS_BASE | 0x30000) diff --git a/arch/arm/mach-mv78xx0/include/mach/bridge-regs.h b/arch/arm/mach-mv78xx0/include/mach/bridge-regs.h index c64dbb9..eb187e0 100644 --- a/arch/arm/mach-mv78xx0/include/mach/bridge-regs.h +++ b/arch/arm/mach-mv78xx0/include/mach/bridge-regs.h @@ -31,5 +31,6 @@ #define IRQ_MASK_HIGH_OFF 0x0014 #define TIMER_VIRT_BASE (BRIDGE_VIRT_BASE | 0x0300) +#define TIMER_PHYS_BASE (BRIDGE_PHYS_BASE | 0x0300) #endif diff --git a/arch/arm/mach-mv78xx0/include/mach/mv78xx0.h b/arch/arm/mach-mv78xx0/include/mach/mv78xx0.h index 3674497..e807c4c 100644 --- a/arch/arm/mach-mv78xx0/include/mach/mv78xx0.h +++ b/arch/arm/mach-mv78xx0/include/mach/mv78xx0.h @@ -42,6 +42,7 @@ #define MV78XX0_CORE0_REGS_PHYS_BASE 0xf1020000 #define MV78XX0_CORE1_REGS_PHYS_BASE 0xf1024000 #define MV78XX0_CORE_REGS_VIRT_BASE 0xfe400000 +#define MV78XX0_CORE_REGS_PHYS_BASE 0xfe400000 #define MV78XX0_CORE_REGS_SIZE SZ_16K #define MV78XX0_PCIE_IO_PHYS_BASE(i) (0xf0800000 + ((i) << 20)) @@ -59,6 +60,7 @@ * Core-specific peripheral registers. */ #define BRIDGE_VIRT_BASE (MV78XX0_CORE_REGS_VIRT_BASE) +#define BRIDGE_PHYS_BASE (MV78XX0_CORE_REGS_PHYS_BASE) /* * Register Map -- cgit v0.10.2 From 5a081caa0414b9bbb82c17ffab9d6fe66edbb72f Mon Sep 17 00:00:00 2001 From: Ohad Ben-Cohen Date: Wed, 6 Jun 2012 10:09:25 +0300 Subject: rpmsg: avoid premature deallocation of endpoints When an inbound message arrives, the rpmsg core looks up its associated endpoint and invokes the registered callback. If a message arrives while its endpoint is being removed (because the rpmsg driver was removed, or a recovery of a remote processor has kicked in) we must ensure atomicity, i.e.: - Either the ept is removed before it is found or - The ept is found but will not be freed until the callback returns This is achieved by maintaining a per-ept reference count, which, when drops to zero, will trigger deallocation of the ept. With this in hand, it is now forbidden to directly deallocate epts once they have been added to the endpoints idr. Cc: stable Reported-by: Fernando Guzman Lugo Signed-off-by: Ohad Ben-Cohen diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c index 75506ec..9623327 100644 --- a/drivers/rpmsg/virtio_rpmsg_bus.c +++ b/drivers/rpmsg/virtio_rpmsg_bus.c @@ -188,6 +188,26 @@ static int rpmsg_uevent(struct device *dev, struct kobj_uevent_env *env) rpdev->id.name); } +/** + * __ept_release() - deallocate an rpmsg endpoint + * @kref: the ept's reference count + * + * This function deallocates an ept, and is invoked when its @kref refcount + * drops to zero. + * + * Never invoke this function directly! + */ +static void __ept_release(struct kref *kref) +{ + struct rpmsg_endpoint *ept = container_of(kref, struct rpmsg_endpoint, + refcount); + /* + * At this point no one holds a reference to ept anymore, + * so we can directly free it + */ + kfree(ept); +} + /* for more info, see below documentation of rpmsg_create_ept() */ static struct rpmsg_endpoint *__rpmsg_create_ept(struct virtproc_info *vrp, struct rpmsg_channel *rpdev, rpmsg_rx_cb_t cb, @@ -206,6 +226,8 @@ static struct rpmsg_endpoint *__rpmsg_create_ept(struct virtproc_info *vrp, return NULL; } + kref_init(&ept->refcount); + ept->rpdev = rpdev; ept->cb = cb; ept->priv = priv; @@ -238,7 +260,7 @@ rem_idr: idr_remove(&vrp->endpoints, request); free_ept: mutex_unlock(&vrp->endpoints_lock); - kfree(ept); + kref_put(&ept->refcount, __ept_release); return NULL; } @@ -306,7 +328,7 @@ __rpmsg_destroy_ept(struct virtproc_info *vrp, struct rpmsg_endpoint *ept) idr_remove(&vrp->endpoints, ept->addr); mutex_unlock(&vrp->endpoints_lock); - kfree(ept); + kref_put(&ept->refcount, __ept_release); } /** @@ -790,7 +812,13 @@ static void rpmsg_recv_done(struct virtqueue *rvq) /* use the dst addr to fetch the callback of the appropriate user */ mutex_lock(&vrp->endpoints_lock); + ept = idr_find(&vrp->endpoints, msg->dst); + + /* let's make sure no one deallocates ept while we use it */ + if (ept) + kref_get(&ept->refcount); + mutex_unlock(&vrp->endpoints_lock); if (ept && ept->cb) @@ -798,6 +826,10 @@ static void rpmsg_recv_done(struct virtqueue *rvq) else dev_warn(dev, "msg received with no recepient\n"); + /* farewell, ept, we don't need you anymore */ + if (ept) + kref_put(&ept->refcount, __ept_release); + /* publish the real size of the buffer */ sg_init_one(&sg, msg, RPMSG_BUF_SIZE); diff --git a/include/linux/rpmsg.h b/include/linux/rpmsg.h index a8e50e4..195f373 100644 --- a/include/linux/rpmsg.h +++ b/include/linux/rpmsg.h @@ -38,6 +38,7 @@ #include #include #include +#include /* The feature bitmap for virtio rpmsg */ #define VIRTIO_RPMSG_F_NS 0 /* RP supports name service notifications */ @@ -120,6 +121,7 @@ typedef void (*rpmsg_rx_cb_t)(struct rpmsg_channel *, void *, int, void *, u32); /** * struct rpmsg_endpoint - binds a local rpmsg address to its user * @rpdev: rpmsg channel device + * @refcount: when this drops to zero, the ept is deallocated * @cb: rx callback handler * @addr: local rpmsg address * @priv: private data for the driver's use @@ -140,6 +142,7 @@ typedef void (*rpmsg_rx_cb_t)(struct rpmsg_channel *, void *, int, void *, u32); */ struct rpmsg_endpoint { struct rpmsg_channel *rpdev; + struct kref refcount; rpmsg_rx_cb_t cb; u32 addr; void *priv; -- cgit v0.10.2 From 15fd943af50dbc5f7f4de33835795c72595f7bf4 Mon Sep 17 00:00:00 2001 From: Ohad Ben-Cohen Date: Thu, 7 Jun 2012 15:39:35 +0300 Subject: rpmsg: make sure inflight messages don't invoke just-removed callbacks When inbound messages arrive, rpmsg core looks up their associated endpoint (by destination address) and then invokes their callback. We've made sure that endpoints will never be de-allocated after they were found by rpmsg core, but we also need to protect against the (rare) scenario where the rpmsg driver was just removed, and its callback function isn't available anymore. This is achieved by introducing a callback mutex, which must be taken before the callback is invoked, and, obviously, before it is removed. Cc: stable Reported-by: Fernando Guzman Lugo Signed-off-by: Ohad Ben-Cohen diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c index 9623327..39d3aa4 100644 --- a/drivers/rpmsg/virtio_rpmsg_bus.c +++ b/drivers/rpmsg/virtio_rpmsg_bus.c @@ -227,6 +227,7 @@ static struct rpmsg_endpoint *__rpmsg_create_ept(struct virtproc_info *vrp, } kref_init(&ept->refcount); + mutex_init(&ept->cb_lock); ept->rpdev = rpdev; ept->cb = cb; @@ -324,10 +325,16 @@ EXPORT_SYMBOL(rpmsg_create_ept); static void __rpmsg_destroy_ept(struct virtproc_info *vrp, struct rpmsg_endpoint *ept) { + /* make sure new inbound messages can't find this ept anymore */ mutex_lock(&vrp->endpoints_lock); idr_remove(&vrp->endpoints, ept->addr); mutex_unlock(&vrp->endpoints_lock); + /* make sure in-flight inbound messages won't invoke cb anymore */ + mutex_lock(&ept->cb_lock); + ept->cb = NULL; + mutex_unlock(&ept->cb_lock); + kref_put(&ept->refcount, __ept_release); } @@ -821,14 +828,20 @@ static void rpmsg_recv_done(struct virtqueue *rvq) mutex_unlock(&vrp->endpoints_lock); - if (ept && ept->cb) - ept->cb(ept->rpdev, msg->data, msg->len, ept->priv, msg->src); - else - dev_warn(dev, "msg received with no recepient\n"); + if (ept) { + /* make sure ept->cb doesn't go away while we use it */ + mutex_lock(&ept->cb_lock); - /* farewell, ept, we don't need you anymore */ - if (ept) + if (ept->cb) + ept->cb(ept->rpdev, msg->data, msg->len, ept->priv, + msg->src); + + mutex_unlock(&ept->cb_lock); + + /* farewell, ept, we don't need you anymore */ kref_put(&ept->refcount, __ept_release); + } else + dev_warn(dev, "msg received with no recepient\n"); /* publish the real size of the buffer */ sg_init_one(&sg, msg, RPMSG_BUF_SIZE); diff --git a/include/linux/rpmsg.h b/include/linux/rpmsg.h index 195f373..82a6739 100644 --- a/include/linux/rpmsg.h +++ b/include/linux/rpmsg.h @@ -39,6 +39,7 @@ #include #include #include +#include /* The feature bitmap for virtio rpmsg */ #define VIRTIO_RPMSG_F_NS 0 /* RP supports name service notifications */ @@ -123,6 +124,7 @@ typedef void (*rpmsg_rx_cb_t)(struct rpmsg_channel *, void *, int, void *, u32); * @rpdev: rpmsg channel device * @refcount: when this drops to zero, the ept is deallocated * @cb: rx callback handler + * @cb_lock: must be taken before accessing/changing @cb * @addr: local rpmsg address * @priv: private data for the driver's use * @@ -144,6 +146,7 @@ struct rpmsg_endpoint { struct rpmsg_channel *rpdev; struct kref refcount; rpmsg_rx_cb_t cb; + struct mutex cb_lock; u32 addr; void *priv; }; -- cgit v0.10.2 From b0a70cc80e754f5ad7d83c96e005fa1d5cb83514 Mon Sep 17 00:00:00 2001 From: Paul Walmsley Date: Wed, 4 Jul 2012 06:55:29 -0600 Subject: ARM: OMAP4: hwmod data: temporarily comment out data for the usb_host_fs and aess IP blocks MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The OMAP4 usb_host_fs (OHCI) and AESS IP blocks require some special programming for them to enter idle. Without this programming, they will prevent the rest of the chip from entering full chip idle. To implement the idle programming cleanly, this will take some coordination between maintainers. This is likely to take some time, so it is probably best to leave this for 3.6 or 3.7. So, in the meantime, prevent these IP blocks from being registered. Later, once the appropriate support is available, this patch can be reverted. This second version comments out the IP block data since Benoît didn't like removing it. Signed-off-by: Paul Walmsley Cc: Benoît Cousson diff --git a/arch/arm/mach-omap2/omap_hwmod_44xx_data.c b/arch/arm/mach-omap2/omap_hwmod_44xx_data.c index f30e861..0a1b247 100644 --- a/arch/arm/mach-omap2/omap_hwmod_44xx_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_44xx_data.c @@ -3864,7 +3864,7 @@ static struct omap_hwmod_ocp_if omap44xx_l4_cfg__l3_main_2 = { }; /* usb_host_fs -> l3_main_2 */ -static struct omap_hwmod_ocp_if omap44xx_usb_host_fs__l3_main_2 = { +static struct omap_hwmod_ocp_if __maybe_unused omap44xx_usb_host_fs__l3_main_2 = { .master = &omap44xx_usb_host_fs_hwmod, .slave = &omap44xx_l3_main_2_hwmod, .clk = "l3_div_ck", @@ -3922,7 +3922,7 @@ static struct omap_hwmod_ocp_if omap44xx_l4_cfg__l3_main_3 = { }; /* aess -> l4_abe */ -static struct omap_hwmod_ocp_if omap44xx_aess__l4_abe = { +static struct omap_hwmod_ocp_if __maybe_unused omap44xx_aess__l4_abe = { .master = &omap44xx_aess_hwmod, .slave = &omap44xx_l4_abe_hwmod, .clk = "ocp_abe_iclk", @@ -4013,7 +4013,7 @@ static struct omap_hwmod_addr_space omap44xx_aess_addrs[] = { }; /* l4_abe -> aess */ -static struct omap_hwmod_ocp_if omap44xx_l4_abe__aess = { +static struct omap_hwmod_ocp_if __maybe_unused omap44xx_l4_abe__aess = { .master = &omap44xx_l4_abe_hwmod, .slave = &omap44xx_aess_hwmod, .clk = "ocp_abe_iclk", @@ -4031,7 +4031,7 @@ static struct omap_hwmod_addr_space omap44xx_aess_dma_addrs[] = { }; /* l4_abe -> aess (dma) */ -static struct omap_hwmod_ocp_if omap44xx_l4_abe__aess_dma = { +static struct omap_hwmod_ocp_if __maybe_unused omap44xx_l4_abe__aess_dma = { .master = &omap44xx_l4_abe_hwmod, .slave = &omap44xx_aess_hwmod, .clk = "ocp_abe_iclk", @@ -5857,7 +5857,7 @@ static struct omap_hwmod_addr_space omap44xx_usb_host_fs_addrs[] = { }; /* l4_cfg -> usb_host_fs */ -static struct omap_hwmod_ocp_if omap44xx_l4_cfg__usb_host_fs = { +static struct omap_hwmod_ocp_if __maybe_unused omap44xx_l4_cfg__usb_host_fs = { .master = &omap44xx_l4_cfg_hwmod, .slave = &omap44xx_usb_host_fs_hwmod, .clk = "l4_div_ck", @@ -6014,13 +6014,13 @@ static struct omap_hwmod_ocp_if *omap44xx_hwmod_ocp_ifs[] __initdata = { &omap44xx_iva__l3_main_2, &omap44xx_l3_main_1__l3_main_2, &omap44xx_l4_cfg__l3_main_2, - &omap44xx_usb_host_fs__l3_main_2, + /* &omap44xx_usb_host_fs__l3_main_2, */ &omap44xx_usb_host_hs__l3_main_2, &omap44xx_usb_otg_hs__l3_main_2, &omap44xx_l3_main_1__l3_main_3, &omap44xx_l3_main_2__l3_main_3, &omap44xx_l4_cfg__l3_main_3, - &omap44xx_aess__l4_abe, + /* &omap44xx_aess__l4_abe, */ &omap44xx_dsp__l4_abe, &omap44xx_l3_main_1__l4_abe, &omap44xx_mpu__l4_abe, @@ -6029,8 +6029,8 @@ static struct omap_hwmod_ocp_if *omap44xx_hwmod_ocp_ifs[] __initdata = { &omap44xx_l4_cfg__l4_wkup, &omap44xx_mpu__mpu_private, &omap44xx_l4_cfg__ocp_wp_noc, - &omap44xx_l4_abe__aess, - &omap44xx_l4_abe__aess_dma, + /* &omap44xx_l4_abe__aess, */ + /* &omap44xx_l4_abe__aess_dma, */ &omap44xx_l3_main_2__c2c, &omap44xx_l4_wkup__counter_32k, &omap44xx_l4_cfg__ctrl_module_core, @@ -6136,7 +6136,7 @@ static struct omap_hwmod_ocp_if *omap44xx_hwmod_ocp_ifs[] __initdata = { &omap44xx_l4_per__uart2, &omap44xx_l4_per__uart3, &omap44xx_l4_per__uart4, - &omap44xx_l4_cfg__usb_host_fs, + /* &omap44xx_l4_cfg__usb_host_fs, */ &omap44xx_l4_cfg__usb_host_hs, &omap44xx_l4_cfg__usb_otg_hs, &omap44xx_l4_cfg__usb_tll_hs, -- cgit v0.10.2 From d7a0b5133f6b1d53d693b9b9873e64e3c8f0db0e Mon Sep 17 00:00:00 2001 From: Benoit Cousson Date: Wed, 4 Jul 2012 06:55:29 -0600 Subject: ARM: OMAP2+: hwmod data: Fix wrong McBSP clock alias on OMAP4 The commit 503d0ea24d1d3dd3db95e5e0edd693da7a2a23eb ARM: OMAP4: hwmod data: Add aliases for McBSP fclk clocks added a wrong "prcm_clk" alias for PRCM clock whereas the McBSP driver and previous OMAPs are using "prcm_fck". It thus lead to the following warning. [ 47.409729] omap-mcbsp: clks: could not clk_get() prcm_fck Fix that by changing the opt_clk role to prcm_fck. Reported-by: Misael Lopez Cruz Signed-off-by: Benoit Cousson Cc: Peter Ujfalusi Tested-by: Sebastien Guiriec Signed-off-by: Paul Walmsley diff --git a/arch/arm/mach-omap2/omap_hwmod_44xx_data.c b/arch/arm/mach-omap2/omap_hwmod_44xx_data.c index 0a1b247..b7bcba5 100644 --- a/arch/arm/mach-omap2/omap_hwmod_44xx_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_44xx_data.c @@ -1928,7 +1928,7 @@ static struct omap_hwmod_dma_info omap44xx_mcbsp1_sdma_reqs[] = { static struct omap_hwmod_opt_clk mcbsp1_opt_clks[] = { { .role = "pad_fck", .clk = "pad_clks_ck" }, - { .role = "prcm_clk", .clk = "mcbsp1_sync_mux_ck" }, + { .role = "prcm_fck", .clk = "mcbsp1_sync_mux_ck" }, }; static struct omap_hwmod omap44xx_mcbsp1_hwmod = { @@ -1963,7 +1963,7 @@ static struct omap_hwmod_dma_info omap44xx_mcbsp2_sdma_reqs[] = { static struct omap_hwmod_opt_clk mcbsp2_opt_clks[] = { { .role = "pad_fck", .clk = "pad_clks_ck" }, - { .role = "prcm_clk", .clk = "mcbsp2_sync_mux_ck" }, + { .role = "prcm_fck", .clk = "mcbsp2_sync_mux_ck" }, }; static struct omap_hwmod omap44xx_mcbsp2_hwmod = { @@ -1998,7 +1998,7 @@ static struct omap_hwmod_dma_info omap44xx_mcbsp3_sdma_reqs[] = { static struct omap_hwmod_opt_clk mcbsp3_opt_clks[] = { { .role = "pad_fck", .clk = "pad_clks_ck" }, - { .role = "prcm_clk", .clk = "mcbsp3_sync_mux_ck" }, + { .role = "prcm_fck", .clk = "mcbsp3_sync_mux_ck" }, }; static struct omap_hwmod omap44xx_mcbsp3_hwmod = { @@ -2033,7 +2033,7 @@ static struct omap_hwmod_dma_info omap44xx_mcbsp4_sdma_reqs[] = { static struct omap_hwmod_opt_clk mcbsp4_opt_clks[] = { { .role = "pad_fck", .clk = "pad_clks_ck" }, - { .role = "prcm_clk", .clk = "mcbsp4_sync_mux_ck" }, + { .role = "prcm_fck", .clk = "mcbsp4_sync_mux_ck" }, }; static struct omap_hwmod omap44xx_mcbsp4_hwmod = { -- cgit v0.10.2 From 0c47935c5b5cd4916cf1c1ed4a2894807f7bcc3e Mon Sep 17 00:00:00 2001 From: Daniel Nicoletti Date: Wed, 4 Jul 2012 10:20:31 -0300 Subject: HID: add battery quirk for Apple Wireless ANSI Add USB_DEVICE_ID_APPLE_ALU_WIRELESS_ANSI, to the quirk list since it report wrong feature type and wrong percentage range. Signed-off-by: Daniel Nicoletti Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-input.c b/drivers/hid/hid-input.c index 132b001..5301006 100644 --- a/drivers/hid/hid-input.c +++ b/drivers/hid/hid-input.c @@ -301,6 +301,9 @@ static const struct hid_device_id hid_battery_quirks[] = { { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_ALU_WIRELESS_2011_ANSI), HID_BATTERY_QUIRK_PERCENT | HID_BATTERY_QUIRK_FEATURE }, + { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_APPLE, + USB_DEVICE_ID_APPLE_ALU_WIRELESS_ANSI), + HID_BATTERY_QUIRK_PERCENT | HID_BATTERY_QUIRK_FEATURE }, {} }; -- cgit v0.10.2 From 32e1eb59f7042c03cb8dbb598e7c97fddbb62ac2 Mon Sep 17 00:00:00 2001 From: Russell King Date: Wed, 4 Jul 2012 17:04:57 +0100 Subject: ARM: fix mach-versatile/pci.c warning arch/arm/mach-versatile/pci.c: In function 'versatile_map_irq': arch/arm/mach-versatile/pci.c:342: warning: unused variable 'devslot' Signed-off-by: Russell King diff --git a/arch/arm/mach-versatile/pci.c b/arch/arm/mach-versatile/pci.c index bec933b..e95bf84 100644 --- a/arch/arm/mach-versatile/pci.c +++ b/arch/arm/mach-versatile/pci.c @@ -339,7 +339,6 @@ void __init pci_versatile_preinit(void) static int __init versatile_map_irq(const struct pci_dev *dev, u8 slot, u8 pin) { int irq; - int devslot = PCI_SLOT(dev->devfn); /* slot, pin, irq * 24 1 27 -- cgit v0.10.2 From 9b7e31bbf4bb58b12e11a7f24b7c3e48bbd2f4da Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Wed, 4 Jul 2012 13:02:56 -0700 Subject: Input: request threaded-only IRQs with IRQF_ONESHOT Since commit 1c6c69525b ("genirq: Reject bogus threaded irq requests") threaded IRQs without a primary handler need to be requested with IRQF_ONESHOT, otherwise the request will fail. This patch adds the IRQF_ONESHOT to input drivers where it is missing. Not modified by this patch are those drivers where the requested IRQ will always be a nested IRQ (e.g. because it's part of an MFD), since for this special case IRQF_ONESHOT is not required to be specified when requesting the IRQ. Signed-off-by: Lars-Peter Clausen Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/joystick/as5011.c b/drivers/input/joystick/as5011.c index 3063464..feeefcb 100644 --- a/drivers/input/joystick/as5011.c +++ b/drivers/input/joystick/as5011.c @@ -281,7 +281,8 @@ static int __devinit as5011_probe(struct i2c_client *client, error = request_threaded_irq(as5011->button_irq, NULL, as5011_button_interrupt, - IRQF_TRIGGER_RISING | IRQF_TRIGGER_FALLING, + IRQF_TRIGGER_RISING | + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, "as5011_button", as5011); if (error < 0) { dev_err(&client->dev, @@ -295,7 +296,7 @@ static int __devinit as5011_probe(struct i2c_client *client, error = request_threaded_irq(as5011->axis_irq, NULL, as5011_axis_interrupt, - plat_data->axis_irqflags, + plat_data->axis_irqflags | IRQF_ONESHOT, "as5011_joystick", as5011); if (error) { dev_err(&client->dev, diff --git a/drivers/input/keyboard/mcs_touchkey.c b/drivers/input/keyboard/mcs_touchkey.c index 64a0ca4..0d77f6c 100644 --- a/drivers/input/keyboard/mcs_touchkey.c +++ b/drivers/input/keyboard/mcs_touchkey.c @@ -178,7 +178,8 @@ static int __devinit mcs_touchkey_probe(struct i2c_client *client, } error = request_threaded_irq(client->irq, NULL, mcs_touchkey_interrupt, - IRQF_TRIGGER_FALLING, client->dev.driver->name, data); + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, + client->dev.driver->name, data); if (error) { dev_err(&client->dev, "Failed to register interrupt\n"); goto err_free_mem; diff --git a/drivers/input/keyboard/mpr121_touchkey.c b/drivers/input/keyboard/mpr121_touchkey.c index caa218a..7613f1c 100644 --- a/drivers/input/keyboard/mpr121_touchkey.c +++ b/drivers/input/keyboard/mpr121_touchkey.c @@ -248,7 +248,7 @@ static int __devinit mpr_touchkey_probe(struct i2c_client *client, error = request_threaded_irq(client->irq, NULL, mpr_touchkey_interrupt, - IRQF_TRIGGER_FALLING, + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, client->dev.driver->name, mpr121); if (error) { dev_err(&client->dev, "Failed to register interrupt\n"); diff --git a/drivers/input/keyboard/qt1070.c b/drivers/input/keyboard/qt1070.c index 0b7b2f8..ca68f29 100644 --- a/drivers/input/keyboard/qt1070.c +++ b/drivers/input/keyboard/qt1070.c @@ -201,7 +201,8 @@ static int __devinit qt1070_probe(struct i2c_client *client, msleep(QT1070_RESET_TIME); err = request_threaded_irq(client->irq, NULL, qt1070_interrupt, - IRQF_TRIGGER_NONE, client->dev.driver->name, data); + IRQF_TRIGGER_NONE | IRQF_ONESHOT, + client->dev.driver->name, data); if (err) { dev_err(&client->dev, "fail to request irq\n"); goto err_free_mem; diff --git a/drivers/input/keyboard/tca6416-keypad.c b/drivers/input/keyboard/tca6416-keypad.c index 3afea3f..c355cdd 100644 --- a/drivers/input/keyboard/tca6416-keypad.c +++ b/drivers/input/keyboard/tca6416-keypad.c @@ -278,7 +278,8 @@ static int __devinit tca6416_keypad_probe(struct i2c_client *client, error = request_threaded_irq(chip->irqnum, NULL, tca6416_keys_isr, - IRQF_TRIGGER_FALLING, + IRQF_TRIGGER_FALLING | + IRQF_ONESHOT, "tca6416-keypad", chip); if (error) { dev_dbg(&client->dev, diff --git a/drivers/input/keyboard/tca8418_keypad.c b/drivers/input/keyboard/tca8418_keypad.c index 5f87b28..893869b 100644 --- a/drivers/input/keyboard/tca8418_keypad.c +++ b/drivers/input/keyboard/tca8418_keypad.c @@ -360,7 +360,7 @@ static int __devinit tca8418_keypad_probe(struct i2c_client *client, client->irq = gpio_to_irq(client->irq); error = request_threaded_irq(client->irq, NULL, tca8418_irq_handler, - IRQF_TRIGGER_FALLING, + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, client->name, keypad_data); if (error) { dev_dbg(&client->dev, diff --git a/drivers/input/keyboard/tnetv107x-keypad.c b/drivers/input/keyboard/tnetv107x-keypad.c index a4a445f..4c34f21 100644 --- a/drivers/input/keyboard/tnetv107x-keypad.c +++ b/drivers/input/keyboard/tnetv107x-keypad.c @@ -227,15 +227,15 @@ static int __devinit keypad_probe(struct platform_device *pdev) goto error_clk; } - error = request_threaded_irq(kp->irq_press, NULL, keypad_irq, 0, - dev_name(dev), kp); + error = request_threaded_irq(kp->irq_press, NULL, keypad_irq, + IRQF_ONESHOT, dev_name(dev), kp); if (error < 0) { dev_err(kp->dev, "Could not allocate keypad press key irq\n"); goto error_irq_press; } - error = request_threaded_irq(kp->irq_release, NULL, keypad_irq, 0, - dev_name(dev), kp); + error = request_threaded_irq(kp->irq_release, NULL, keypad_irq, + IRQF_ONESHOT, dev_name(dev), kp); if (error < 0) { dev_err(kp->dev, "Could not allocate keypad release key irq\n"); goto error_irq_release; diff --git a/drivers/input/misc/ad714x.c b/drivers/input/misc/ad714x.c index 0ac75bb..2e5d5e1 100644 --- a/drivers/input/misc/ad714x.c +++ b/drivers/input/misc/ad714x.c @@ -972,6 +972,7 @@ struct ad714x_chip *ad714x_probe(struct device *dev, u16 bus_type, int irq, struct ad714x_platform_data *plat_data = dev->platform_data; struct ad714x_chip *ad714x; void *drv_mem; + unsigned long irqflags; struct ad714x_button_drv *bt_drv; struct ad714x_slider_drv *sd_drv; @@ -1162,10 +1163,11 @@ struct ad714x_chip *ad714x_probe(struct device *dev, u16 bus_type, int irq, alloc_idx++; } + irqflags = plat_data->irqflags ?: IRQF_TRIGGER_FALLING; + irqflags |= IRQF_ONESHOT; + error = request_threaded_irq(ad714x->irq, NULL, ad714x_interrupt_thread, - plat_data->irqflags ? - plat_data->irqflags : IRQF_TRIGGER_FALLING, - "ad714x_captouch", ad714x); + irqflags, "ad714x_captouch", ad714x); if (error) { dev_err(dev, "can't allocate irq %d\n", ad714x->irq); goto err_unreg_dev; diff --git a/drivers/input/misc/dm355evm_keys.c b/drivers/input/misc/dm355evm_keys.c index 35083c6..c1313d8 100644 --- a/drivers/input/misc/dm355evm_keys.c +++ b/drivers/input/misc/dm355evm_keys.c @@ -213,7 +213,8 @@ static int __devinit dm355evm_keys_probe(struct platform_device *pdev) /* REVISIT: flush the event queue? */ status = request_threaded_irq(keys->irq, NULL, dm355evm_keys_irq, - IRQF_TRIGGER_FALLING, dev_name(&pdev->dev), keys); + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, + dev_name(&pdev->dev), keys); if (status < 0) goto fail2; diff --git a/drivers/input/touchscreen/ad7879.c b/drivers/input/touchscreen/ad7879.c index e2482b4..bd4eb42 100644 --- a/drivers/input/touchscreen/ad7879.c +++ b/drivers/input/touchscreen/ad7879.c @@ -597,7 +597,7 @@ struct ad7879 *ad7879_probe(struct device *dev, u8 devid, unsigned int irq, AD7879_TMR(ts->pen_down_acc_interval); err = request_threaded_irq(ts->irq, NULL, ad7879_irq, - IRQF_TRIGGER_FALLING, + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, dev_name(dev), ts); if (err) { dev_err(dev, "irq %d busy?\n", ts->irq); diff --git a/drivers/input/touchscreen/atmel_mxt_ts.c b/drivers/input/touchscreen/atmel_mxt_ts.c index 42e6450..25fd056 100644 --- a/drivers/input/touchscreen/atmel_mxt_ts.c +++ b/drivers/input/touchscreen/atmel_mxt_ts.c @@ -1149,7 +1149,8 @@ static int __devinit mxt_probe(struct i2c_client *client, goto err_free_object; error = request_threaded_irq(client->irq, NULL, mxt_interrupt, - pdata->irqflags, client->dev.driver->name, data); + pdata->irqflags | IRQF_ONESHOT, + client->dev.driver->name, data); if (error) { dev_err(&client->dev, "Failed to register interrupt\n"); goto err_free_object; diff --git a/drivers/input/touchscreen/bu21013_ts.c b/drivers/input/touchscreen/bu21013_ts.c index f2d03c0..5c487d2 100644 --- a/drivers/input/touchscreen/bu21013_ts.c +++ b/drivers/input/touchscreen/bu21013_ts.c @@ -509,7 +509,8 @@ static int __devinit bu21013_probe(struct i2c_client *client, input_set_drvdata(in_dev, bu21013_data); error = request_threaded_irq(pdata->irq, NULL, bu21013_gpio_irq, - IRQF_TRIGGER_FALLING | IRQF_SHARED, + IRQF_TRIGGER_FALLING | IRQF_SHARED | + IRQF_ONESHOT, DRIVER_TP, bu21013_data); if (error) { dev_err(&client->dev, "request irq %d failed\n", pdata->irq); diff --git a/drivers/input/touchscreen/cy8ctmg110_ts.c b/drivers/input/touchscreen/cy8ctmg110_ts.c index 237753a..464f1bf 100644 --- a/drivers/input/touchscreen/cy8ctmg110_ts.c +++ b/drivers/input/touchscreen/cy8ctmg110_ts.c @@ -251,7 +251,8 @@ static int __devinit cy8ctmg110_probe(struct i2c_client *client, } err = request_threaded_irq(client->irq, NULL, cy8ctmg110_irq_thread, - IRQF_TRIGGER_RISING, "touch_reset_key", ts); + IRQF_TRIGGER_RISING | IRQF_ONESHOT, + "touch_reset_key", ts); if (err < 0) { dev_err(&client->dev, "irq %d busy? error %d\n", client->irq, err); diff --git a/drivers/input/touchscreen/intel-mid-touch.c b/drivers/input/touchscreen/intel-mid-touch.c index 3cd7a83..cf29937 100644 --- a/drivers/input/touchscreen/intel-mid-touch.c +++ b/drivers/input/touchscreen/intel-mid-touch.c @@ -620,7 +620,7 @@ static int __devinit mrstouch_probe(struct platform_device *pdev) MRST_PRESSURE_MIN, MRST_PRESSURE_MAX, 0, 0); err = request_threaded_irq(tsdev->irq, NULL, mrstouch_pendet_irq, - 0, "mrstouch", tsdev); + IRQF_ONESHOT, "mrstouch", tsdev); if (err) { dev_err(tsdev->dev, "unable to allocate irq\n"); goto err_free_mem; diff --git a/drivers/input/touchscreen/pixcir_i2c_ts.c b/drivers/input/touchscreen/pixcir_i2c_ts.c index 72f6ba3..953b4c1 100644 --- a/drivers/input/touchscreen/pixcir_i2c_ts.c +++ b/drivers/input/touchscreen/pixcir_i2c_ts.c @@ -165,7 +165,7 @@ static int __devinit pixcir_i2c_ts_probe(struct i2c_client *client, input_set_drvdata(input, tsdata); error = request_threaded_irq(client->irq, NULL, pixcir_ts_isr, - IRQF_TRIGGER_FALLING, + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, client->name, tsdata); if (error) { dev_err(&client->dev, "Unable to request touchscreen IRQ.\n"); diff --git a/drivers/input/touchscreen/tnetv107x-ts.c b/drivers/input/touchscreen/tnetv107x-ts.c index 7e74880..368d2c6c 100644 --- a/drivers/input/touchscreen/tnetv107x-ts.c +++ b/drivers/input/touchscreen/tnetv107x-ts.c @@ -297,7 +297,7 @@ static int __devinit tsc_probe(struct platform_device *pdev) goto error_clk; } - error = request_threaded_irq(ts->tsc_irq, NULL, tsc_irq, 0, + error = request_threaded_irq(ts->tsc_irq, NULL, tsc_irq, IRQF_ONESHOT, dev_name(dev), ts); if (error < 0) { dev_err(ts->dev, "Could not allocate ts irq\n"); diff --git a/drivers/input/touchscreen/tsc2005.c b/drivers/input/touchscreen/tsc2005.c index b6adeae..5ce3fa8 100644 --- a/drivers/input/touchscreen/tsc2005.c +++ b/drivers/input/touchscreen/tsc2005.c @@ -650,7 +650,8 @@ static int __devinit tsc2005_probe(struct spi_device *spi) tsc2005_stop_scan(ts); error = request_threaded_irq(spi->irq, NULL, tsc2005_irq_thread, - IRQF_TRIGGER_RISING, "tsc2005", ts); + IRQF_TRIGGER_RISING | IRQF_ONESHOT, + "tsc2005", ts); if (error) { dev_err(&spi->dev, "Failed to request irq, err: %d\n", error); goto err_free_mem; -- cgit v0.10.2 From c46d2916f6c03d3c8a2c38148af2e45cdadaf61f Mon Sep 17 00:00:00 2001 From: Lauri Hintsala Date: Thu, 5 Jul 2012 10:31:36 +0300 Subject: ARM: apx4devkit: fix FEC enabling PHY clock Ethernet stopped to work after mxs clk framework change. Signed-off-by: Lauri Hintsala Signed-off-by: Shawn Guo diff --git a/arch/arm/mach-mxs/mach-apx4devkit.c b/arch/arm/mach-mxs/mach-apx4devkit.c index 5e90b9d..f5f0617 100644 --- a/arch/arm/mach-mxs/mach-apx4devkit.c +++ b/arch/arm/mach-mxs/mach-apx4devkit.c @@ -205,6 +205,16 @@ static int apx4devkit_phy_fixup(struct phy_device *phy) return 0; } +static void __init apx4devkit_fec_phy_clk_enable(void) +{ + struct clk *clk; + + /* Enable fec phy clock */ + clk = clk_get_sys("enet_out", NULL); + if (!IS_ERR(clk)) + clk_prepare_enable(clk); +} + static void __init apx4devkit_init(void) { mx28_soc_init(); @@ -225,6 +235,7 @@ static void __init apx4devkit_init(void) phy_register_fixup_for_uid(PHY_ID_KS8051, MICREL_PHY_ID_MASK, apx4devkit_phy_fixup); + apx4devkit_fec_phy_clk_enable(); mx28_add_fec(0, &mx28_fec_pdata); mx28_add_mxs_mmc(0, &apx4devkit_mmc_pdata); -- cgit v0.10.2 From 2dfd06036ba7ae8e7be2daf5a2fff1dac42390bf Mon Sep 17 00:00:00 2001 From: Junxiao Bi Date: Wed, 27 Jun 2012 17:09:54 +0800 Subject: aio: make kiocb->private NUll in init_sync_kiocb() Ocfs2 uses kiocb.*private as a flag of unsigned long size. In commit a11f7e6 ocfs2: serialize unaligned aio, the unaligned io flag is involved in it to serialize the unaligned aio. As *private is not initialized in init_sync_kiocb() of do_sync_write(), this unaligned io flag may be unexpectly set in an aligned dio. And this will cause OCFS2_I(inode)->ip_unaligned_aio decreased to -1 in ocfs2_dio_end_io(), thus the following unaligned dio will hang forever at ocfs2_aiodio_wait() in ocfs2_file_aio_write(). Signed-off-by: Junxiao Bi Cc: stable@vger.kernel.org Acked-by: Jeff Moyer Signed-off-by: Joel Becker diff --git a/include/linux/aio.h b/include/linux/aio.h index 2314ad8..b1a520e 100644 --- a/include/linux/aio.h +++ b/include/linux/aio.h @@ -140,6 +140,7 @@ struct kiocb { (x)->ki_dtor = NULL; \ (x)->ki_obj.tsk = tsk; \ (x)->ki_user_data = 0; \ + (x)->private = NULL; \ } while (0) #define AIO_RING_MAGIC 0xa10a10a1 -- cgit v0.10.2 From 82401bf105495c593544375b4748f48fce70d9c4 Mon Sep 17 00:00:00 2001 From: Russell King Date: Wed, 4 Jul 2012 17:05:28 +0100 Subject: ARM: fix set_domain() macro Avoid polluting drivers with a set_domain() macro, which interferes with structure member names: drivers/net/wireless/ath/ath9k/dfs_pattern_detector.c:294:33: error: macro "set_domain" passed 2 arguments, but takes just 1 Signed-off-by: Russell King diff --git a/arch/arm/include/asm/domain.h b/arch/arm/include/asm/domain.h index 3d22204..6ddbe44 100644 --- a/arch/arm/include/asm/domain.h +++ b/arch/arm/include/asm/domain.h @@ -60,13 +60,13 @@ #ifndef __ASSEMBLY__ #ifdef CONFIG_CPU_USE_DOMAINS -#define set_domain(x) \ - do { \ - __asm__ __volatile__( \ - "mcr p15, 0, %0, c3, c0 @ set domain" \ - : : "r" (x)); \ - isb(); \ - } while (0) +static inline void set_domain(unsigned val) +{ + asm volatile( + "mcr p15, 0, %0, c3, c0 @ set domain" + : : "r" (val)); + isb(); +} #define modify_domain(dom,type) \ do { \ @@ -78,8 +78,8 @@ } while (0) #else -#define set_domain(x) do { } while (0) -#define modify_domain(dom,type) do { } while (0) +static inline void set_domain(unsigned val) { } +static inline void modify_domain(unsigned dom, unsigned type) { } #endif /* -- cgit v0.10.2 From 3b0c06226783ffc836217eb34f7eca311b1e63f7 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Wed, 4 Jul 2012 18:16:30 +0100 Subject: ARM: 7442/1: Revert "remove unused restart trampoline" This reverts commit fa18484d0947b976a769d15c83c50617493c81c1. We need the restart trampoline back so that we can revert a related problematic patch 6b5c8045ecc7e726cdaa2a9d9c8e5008050e1252 ("arm: new way of handling ERESTART_RESTARTBLOCK"). Acked-by: Catalin Marinas Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c index fd2392a..6d3bce5 100644 --- a/arch/arm/kernel/signal.c +++ b/arch/arm/kernel/signal.c @@ -27,6 +27,7 @@ */ #define SWI_SYS_SIGRETURN (0xef000000|(__NR_sigreturn)|(__NR_OABI_SYSCALL_BASE)) #define SWI_SYS_RT_SIGRETURN (0xef000000|(__NR_rt_sigreturn)|(__NR_OABI_SYSCALL_BASE)) +#define SWI_SYS_RESTART (0xef000000|__NR_restart_syscall|__NR_OABI_SYSCALL_BASE) /* * With EABI, the syscall number has to be loaded into r7. @@ -47,6 +48,18 @@ const unsigned long sigreturn_codes[7] = { }; /* + * Either we support OABI only, or we have EABI with the OABI + * compat layer enabled. In the later case we don't know if + * user space is EABI or not, and if not we must not clobber r7. + * Always using the OABI syscall solves that issue and works for + * all those cases. + */ +const unsigned long syscall_restart_code[2] = { + SWI_SYS_RESTART, /* swi __NR_restart_syscall */ + 0xe49df004, /* ldr pc, [sp], #4 */ +}; + +/* * atomically swap in the new signal mask, and wait for a signal. */ asmlinkage int sys_sigsuspend(int restart, unsigned long oldmask, old_sigset_t mask) diff --git a/arch/arm/kernel/signal.h b/arch/arm/kernel/signal.h index 5ff067b7..6fcfe83 100644 --- a/arch/arm/kernel/signal.h +++ b/arch/arm/kernel/signal.h @@ -8,5 +8,7 @@ * published by the Free Software Foundation. */ #define KERN_SIGRETURN_CODE (CONFIG_VECTORS_BASE + 0x00000500) +#define KERN_RESTART_CODE (KERN_SIGRETURN_CODE + sizeof(sigreturn_codes)) extern const unsigned long sigreturn_codes[7]; +extern const unsigned long syscall_restart_code[2]; diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index 4928d89..3647170 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -820,6 +820,8 @@ void __init early_trap_init(void *vectors_base) */ memcpy((void *)(vectors + KERN_SIGRETURN_CODE - CONFIG_VECTORS_BASE), sigreturn_codes, sizeof(sigreturn_codes)); + memcpy((void *)(vectors + KERN_RESTART_CODE - CONFIG_VECTORS_BASE), + syscall_restart_code, sizeof(syscall_restart_code)); flush_icache_range(vectors, vectors + PAGE_SIZE); modify_domain(DOMAIN_USER, DOMAIN_CLIENT); -- cgit v0.10.2 From 433e2f307beff8adba241646ce9108544e0c5a03 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Wed, 4 Jul 2012 18:17:16 +0100 Subject: ARM: 7443/1: Revert "new way of handling ERESTART_RESTARTBLOCK" This reverts commit 6b5c8045ecc7e726cdaa2a9d9c8e5008050e1252. Conflicts: arch/arm/kernel/ptrace.c The new syscall restarting code can lead to problems if we take an interrupt in userspace just before restarting the svc instruction. If a signal is delivered when returning from the interrupt, the TIF_SYSCALL_RESTARTSYS will remain set and cause any syscalls executed from the signal handler to be treated as a restart of the previously interrupted system call. This includes the final sigreturn call, meaning that we may fail to exit from the signal context. Furthermore, if a system call made from the signal handler requires a restart via the restart_block, it is possible to clear the thread flag and fail to restart the originally interrupted system call. The right solution to this problem is to perform the restarting in the kernel, avoiding the possibility of handling a further signal before the restart is complete. Since we're almost at -rc6, let's revert the new method for now and aim for in-kernel restarting at a later date. Acked-by: Catalin Marinas Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h index b79f8e9..af7b0bd 100644 --- a/arch/arm/include/asm/thread_info.h +++ b/arch/arm/include/asm/thread_info.h @@ -148,7 +148,6 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ #define TIF_SYSCALL_TRACE 8 #define TIF_SYSCALL_AUDIT 9 -#define TIF_SYSCALL_RESTARTSYS 10 #define TIF_POLLING_NRFLAG 16 #define TIF_USING_IWMMXT 17 #define TIF_MEMDIE 18 /* is terminating due to OOM killer */ @@ -164,11 +163,9 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, #define _TIF_POLLING_NRFLAG (1 << TIF_POLLING_NRFLAG) #define _TIF_USING_IWMMXT (1 << TIF_USING_IWMMXT) #define _TIF_SECCOMP (1 << TIF_SECCOMP) -#define _TIF_SYSCALL_RESTARTSYS (1 << TIF_SYSCALL_RESTARTSYS) /* Checks for any syscall work in entry-common.S */ -#define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ - _TIF_SYSCALL_RESTARTSYS) +#define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT) /* * Change these and you break ASM code in entry-common.S diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index 5700a7a..14e3826 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -25,7 +25,6 @@ #include #include #include -#include #include #include @@ -918,8 +917,6 @@ asmlinkage int syscall_trace(int why, struct pt_regs *regs, int scno) audit_syscall_entry(AUDIT_ARCH_ARM, scno, regs->ARM_r0, regs->ARM_r1, regs->ARM_r2, regs->ARM_r3); - if (why == 0 && test_and_clear_thread_flag(TIF_SYSCALL_RESTARTSYS)) - scno = __NR_restart_syscall - __NR_SYSCALL_BASE; if (!test_thread_flag(TIF_SYSCALL_TRACE)) return scno; diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c index 6d3bce5..536c5d6 100644 --- a/arch/arm/kernel/signal.c +++ b/arch/arm/kernel/signal.c @@ -605,10 +605,12 @@ static void do_signal(struct pt_regs *regs, int syscall) case -ERESTARTNOHAND: case -ERESTARTSYS: case -ERESTARTNOINTR: - case -ERESTART_RESTARTBLOCK: regs->ARM_r0 = regs->ARM_ORIG_r0; regs->ARM_pc = restart_addr; break; + case -ERESTART_RESTARTBLOCK: + regs->ARM_r0 = -EINTR; + break; } } @@ -624,14 +626,12 @@ static void do_signal(struct pt_regs *regs, int syscall) * debugger has chosen to restart at a different PC. */ if (regs->ARM_pc == restart_addr) { - if (retval == -ERESTARTNOHAND || - retval == -ERESTART_RESTARTBLOCK + if (retval == -ERESTARTNOHAND || (retval == -ERESTARTSYS && !(ka.sa.sa_flags & SA_RESTART))) { regs->ARM_r0 = -EINTR; regs->ARM_pc = continue_addr; } - clear_thread_flag(TIF_SYSCALL_RESTARTSYS); } handle_signal(signr, &ka, &info, regs); @@ -645,8 +645,29 @@ static void do_signal(struct pt_regs *regs, int syscall) * ignore the restart. */ if (retval == -ERESTART_RESTARTBLOCK - && regs->ARM_pc == restart_addr) - set_thread_flag(TIF_SYSCALL_RESTARTSYS); + && regs->ARM_pc == continue_addr) { + if (thumb_mode(regs)) { + regs->ARM_r7 = __NR_restart_syscall - __NR_SYSCALL_BASE; + regs->ARM_pc -= 2; + } else { +#if defined(CONFIG_AEABI) && !defined(CONFIG_OABI_COMPAT) + regs->ARM_r7 = __NR_restart_syscall; + regs->ARM_pc -= 4; +#else + u32 __user *usp; + + regs->ARM_sp -= 4; + usp = (u32 __user *)regs->ARM_sp; + + if (put_user(regs->ARM_pc, usp) == 0) { + regs->ARM_pc = KERN_RESTART_CODE; + } else { + regs->ARM_sp += 4; + force_sigsegv(0, current); + } +#endif + } + } } restore_saved_sigmask(); -- cgit v0.10.2 From fdeb8e35fd59e79dec385f98eb4b6d2e3398264b Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Wed, 4 Jul 2012 18:15:42 +0100 Subject: ARM: 7441/1: perf: return -EOPNOTSUPP if requested mode exclusion is unavailable We currently return -EPERM if the user requests mode exclusion that is not supported by the CPU. This looks pretty confusing from userspace and is inconsistent with other architectures (ppc, x86). This patch returns -EOPNOTSUPP instead. Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index 186c8cb..a02eada 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -503,7 +503,7 @@ __hw_perf_event_init(struct perf_event *event) event_requires_mode_exclusion(&event->attr)) { pr_debug("ARM performance counters do not support " "mode exclusion\n"); - return -EPERM; + return -EOPNOTSUPP; } /* -- cgit v0.10.2 From f8b435bb918412c9410da4c4b0b02b6b3d99b27c Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Wed, 4 Jul 2012 07:37:37 +0100 Subject: ARM: 7440/1: kprobes: only test 'sub pc, pc, #1b-2b+8-2' on ARMv6 'sub pc, pc, #1b-2b+8-2' results in address<1:0> == '10'. sub pc, pc, #const (== ADR pc, #const) performs an interworking branch (BXWritePC()) on ARMv7+ and a simple branch (BranchWritePC()) on earlier versions. In ARM state, BXWritePC() is UNPREDICTABLE when address<1:0> == '10'. In ARM state on ARMv6+, BranchWritePC() ignores address<1:0>. Before ARMv6, BranchWritePC() is UNPREDICTABLE if address<1:0> != '00' So the instruction is UNPREDICTABLE both before and after v6. Acked-by: Jon Medhurst Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/kernel/kprobes-test-arm.c b/arch/arm/kernel/kprobes-test-arm.c index ba32b39..38c1a3b 100644 --- a/arch/arm/kernel/kprobes-test-arm.c +++ b/arch/arm/kernel/kprobes-test-arm.c @@ -187,8 +187,8 @@ void kprobe_arm_test_cases(void) TEST_BF_R ("mov pc, r",0,2f,"") TEST_BF_RR("mov pc, r",0,2f,", asl r",1,0,"") TEST_BB( "sub pc, pc, #1b-2b+8") -#if __LINUX_ARM_ARCH__ >= 6 - TEST_BB( "sub pc, pc, #1b-2b+8-2") /* UNPREDICTABLE before ARMv6 */ +#if __LINUX_ARM_ARCH__ == 6 && !defined(CONFIG_CPU_V7) + TEST_BB( "sub pc, pc, #1b-2b+8-2") /* UNPREDICTABLE before and after ARMv6 */ #endif TEST_BB_R( "sub pc, pc, r",14, 1f-2f+8,"") TEST_BB_R( "rsb pc, r",14,1f-2f+8,", pc") -- cgit v0.10.2 From b89d607b590397c04b63d94a9e2fca9649917955 Mon Sep 17 00:00:00 2001 From: Russell King Date: Thu, 5 Jul 2012 13:06:32 +0100 Subject: ARM: fix warnings about atomic64_read Fix: net/netfilter/xt_connbytes.c: In function 'connbytes_mt': net/netfilter/xt_connbytes.c:43: warning: passing argument 1 of 'atomic64_read' discards qualifiers from pointer target type ... by adding the missing const. Signed-off-by: Russell King diff --git a/arch/arm/include/asm/atomic.h b/arch/arm/include/asm/atomic.h index 68374ba..c79f61f 100644 --- a/arch/arm/include/asm/atomic.h +++ b/arch/arm/include/asm/atomic.h @@ -243,7 +243,7 @@ typedef struct { #define ATOMIC64_INIT(i) { (i) } -static inline u64 atomic64_read(atomic64_t *v) +static inline u64 atomic64_read(const atomic64_t *v) { u64 result; -- cgit v0.10.2 From 09b2ad13da3ac7c717dd86bfca7072d9b36f7449 Mon Sep 17 00:00:00 2001 From: Russell King Date: Thu, 5 Jul 2012 13:11:31 +0100 Subject: ARM: fix warning caused by wrongly typed arm_dma_limit arch/arm/mm/init.c: In function 'arm_memblock_init': arch/arm/mm/init.c:380: warning: comparison of distinct pointer types lacks a cast by fixing the typecast in its definition when DMA_ZONE is disabled. This was missed in 4986e5c7c (ARM: mm: fix type of the arm_dma_limit global variable). Signed-off-by: Russell King diff --git a/arch/arm/mm/mm.h b/arch/arm/mm/mm.h index c471436..2e8a1ef 100644 --- a/arch/arm/mm/mm.h +++ b/arch/arm/mm/mm.h @@ -64,7 +64,7 @@ extern void __flush_dcache_page(struct address_space *mapping, struct page *page #ifdef CONFIG_ZONE_DMA extern phys_addr_t arm_dma_limit; #else -#define arm_dma_limit ((u32)~0) +#define arm_dma_limit ((phys_addr_t)~0) #endif extern phys_addr_t arm_lowmem_limit; -- cgit v0.10.2 From 2e2070c85aa19f6c5bb3642a1429abf42f101e8d Mon Sep 17 00:00:00 2001 From: Alessandro Rubini Date: Sun, 27 May 2012 22:55:41 +0200 Subject: gpio-sta2x11: don't use pdata if null If there is no platform data available, the driver shouldn't use the pointer or it will oops. Since things will mostly work nonetheless, (the BIOS may have set up the pins properly), I'd better not fail the probe even in this case. Signed-off-by: Alessandro Rubini Acked-by: Giancarlo Asnaghi Signed-off-by: Linus Walleij diff --git a/drivers/gpio/gpio-sta2x11.c b/drivers/gpio/gpio-sta2x11.c index 38416be..6064fb3 100644 --- a/drivers/gpio/gpio-sta2x11.c +++ b/drivers/gpio/gpio-sta2x11.c @@ -383,8 +383,9 @@ static int __devinit gsta_probe(struct platform_device *dev) } spin_lock_init(&chip->lock); gsta_gpio_setup(chip); - for (i = 0; i < GSTA_NR_GPIO; i++) - gsta_set_config(chip, i, gpio_pdata->pinconfig[i]); + if (gpio_pdata) + for (i = 0; i < GSTA_NR_GPIO; i++) + gsta_set_config(chip, i, gpio_pdata->pinconfig[i]); /* 384 was used in previous code: be compatible for other drivers */ err = irq_alloc_descs(-1, 384, GSTA_NR_GPIO, NUMA_NO_NODE); -- cgit v0.10.2 From afcc0f8c174ff34305146878c40e7dcc72fcbbdf Mon Sep 17 00:00:00 2001 From: Christian Dietrich Date: Thu, 31 May 2012 13:12:57 +0200 Subject: gpio/msm_v1: CONFIG_GPIO_MSM_V1 is only available on three SoCs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The feature GPIO_MSM_V1 is only available on three SoCs. On all other MSM SoCs the INT_GPIO_GROUP{1,2} is undeclared, but Kconfig does allow such configurations. Therefore the produced configuration is valid, but does not compile. The problem is fixed by adding the missing Kconfig constraints. drivers/gpio/gpio-msm-v1.c: In function ‘msm_init_gpio’: drivers/gpio/gpio-msm-v1.c:629:26: error: 'INT_GPIO_GROUP1' undeclared drivers/gpio/gpio-msm-v1.c:630:26: error: 'INT_GPIO_GROUP2' undeclared Signed-off-by: Christian Dietrich Signed-off-by: Linus Walleij diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig index c4067d0..542f0c0 100644 --- a/drivers/gpio/Kconfig +++ b/drivers/gpio/Kconfig @@ -136,7 +136,7 @@ config GPIO_MPC8XXX config GPIO_MSM_V1 tristate "Qualcomm MSM GPIO v1" - depends on GPIOLIB && ARCH_MSM + depends on GPIOLIB && ARCH_MSM && (ARCH_MSM7X00A || ARCH_MSM7X30 || ARCH_QSD8X50) help Say yes here to support the GPIO interface on ARM v6 based Qualcomm MSM chips. Most of the pins on the MSM can be -- cgit v0.10.2 From 8cd578b6e28693f357867a77598a88ef3deb6b39 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Sat, 9 Jun 2012 11:07:56 +0800 Subject: gpiolib: wm8994: Pay attention to the value set when enabling as output Not paying attention to the value being set is a bad thing because it means that we'll not set the hardware up to reflect what was requested. Not setting the hardware up to reflect what was requested means that the caller won't get the results they wanted. Signed-off-by: Mark Brown Cc: stable@vger.kernel.org Signed-off-by: Linus Walleij diff --git a/drivers/gpio/gpio-wm8994.c b/drivers/gpio/gpio-wm8994.c index 92ea535..aa61ad2 100644 --- a/drivers/gpio/gpio-wm8994.c +++ b/drivers/gpio/gpio-wm8994.c @@ -89,8 +89,11 @@ static int wm8994_gpio_direction_out(struct gpio_chip *chip, struct wm8994_gpio *wm8994_gpio = to_wm8994_gpio(chip); struct wm8994 *wm8994 = wm8994_gpio->wm8994; + if (value) + value = WM8994_GPN_LVL; + return wm8994_set_bits(wm8994, WM8994_GPIO_1 + offset, - WM8994_GPN_DIR, 0); + WM8994_GPN_DIR | WM8994_GPN_LVL, value); } static void wm8994_gpio_set(struct gpio_chip *chip, unsigned offset, int value) -- cgit v0.10.2 From 3996bfc787a3b3d7c7ad97b7519247f9f2ac4068 Mon Sep 17 00:00:00 2001 From: Stephen Warren Date: Thu, 7 Jun 2012 17:56:09 -0600 Subject: gpio: export devm_gpio_request_one Without this, modules can't use this API, leading to build failures. Signed-off-by: Stephen Warren Signed-off-by: Linus Walleij diff --git a/drivers/gpio/devres.c b/drivers/gpio/devres.c index 9e9947c..1077754 100644 --- a/drivers/gpio/devres.c +++ b/drivers/gpio/devres.c @@ -98,6 +98,7 @@ int devm_gpio_request_one(struct device *dev, unsigned gpio, return 0; } +EXPORT_SYMBOL(devm_gpio_request_one); /** * devm_gpio_free - free an interrupt -- cgit v0.10.2 From 42b14cb037fefa33a2ff51c4d3915a49c71de3d5 Mon Sep 17 00:00:00 2001 From: Roland Stigge Date: Mon, 18 Jun 2012 11:28:26 +0200 Subject: mips: pci-lantiq: Fix check for valid gpio This patch fixes two checks for valid gpio number, formerly (wrongly) considering zero as invalid, now using gpio_is_valid(). Signed-off-by: Roland Stigge Signed-off-by: Linus Walleij diff --git a/arch/mips/pci/pci-lantiq.c b/arch/mips/pci/pci-lantiq.c index ea45353..075d87a 100644 --- a/arch/mips/pci/pci-lantiq.c +++ b/arch/mips/pci/pci-lantiq.c @@ -129,7 +129,7 @@ static int __devinit ltq_pci_startup(struct platform_device *pdev) /* setup reset gpio used by pci */ reset_gpio = of_get_named_gpio(node, "gpio-reset", 0); - if (reset_gpio > 0) + if (gpio_is_valid(reset_gpio)) devm_gpio_request(&pdev->dev, reset_gpio, "pci-reset"); /* enable auto-switching between PCI and EBU */ @@ -192,7 +192,7 @@ static int __devinit ltq_pci_startup(struct platform_device *pdev) ltq_ebu_w32(ltq_ebu_r32(LTQ_EBU_PCC_IEN) | 0x10, LTQ_EBU_PCC_IEN); /* toggle reset pin */ - if (reset_gpio > 0) { + if (gpio_is_valid(reset_gpio)) { __gpio_set_value(reset_gpio, 0); wmb(); mdelay(1); -- cgit v0.10.2 From f567fde24640cf6f2d6416196bfc8b3fefc8e433 Mon Sep 17 00:00:00 2001 From: Laxman Dewangan Date: Wed, 20 Jun 2012 14:14:05 +0530 Subject: gpio: fix bits conflict for gpio flags The bit 2 and 3 in GPIO flag are allocated for the flag OPEN_DRAIN/OPEN_SOURCE. These bits are reused for the flag EXPORT/EXPORT_CHANGEABLE and so creating conflict. Fix this conflict by assigning bit 4 and 5 for the flag EXPORT/EXPORT_CHANGEABLE. Signed-off-by: Laxman Dewangan Signed-off-by: Linus Walleij diff --git a/include/linux/gpio.h b/include/linux/gpio.h index f07fc2d..2e31e8b 100644 --- a/include/linux/gpio.h +++ b/include/linux/gpio.h @@ -22,8 +22,8 @@ /* Gpio pin is open source */ #define GPIOF_OPEN_SOURCE (1 << 3) -#define GPIOF_EXPORT (1 << 2) -#define GPIOF_EXPORT_CHANGEABLE (1 << 3) +#define GPIOF_EXPORT (1 << 4) +#define GPIOF_EXPORT_CHANGEABLE (1 << 5) #define GPIOF_EXPORT_DIR_FIXED (GPIOF_EXPORT) #define GPIOF_EXPORT_DIR_CHANGEABLE (GPIOF_EXPORT | GPIOF_EXPORT_CHANGEABLE) -- cgit v0.10.2 From 33a4e985bab25d6753b52d1322b4e2fff706dd4f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Wed, 6 Jun 2012 11:49:23 +0200 Subject: gpio/mxc: make irqs work for fsl,imx21-gpio devices MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The chained handler was set for the platform device with id == 0. When the gpio devices are instantiated by a device tree, all have id == -1 and so the handler was unset resulting in unusable gpio irqs on i.MX21 and i.MX27 (when using oftree). Acked-by: Shawn Guo Cc: Grant Likely Signed-off-by: Uwe Kleine-König Signed-off-by: Linus Walleij diff --git a/drivers/gpio/gpio-mxc.c b/drivers/gpio/gpio-mxc.c index c337143..c89c4c1 100644 --- a/drivers/gpio/gpio-mxc.c +++ b/drivers/gpio/gpio-mxc.c @@ -398,10 +398,12 @@ static int __devinit mxc_gpio_probe(struct platform_device *pdev) writel(~0, port->base + GPIO_ISR); if (mxc_gpio_hwtype == IMX21_GPIO) { - /* setup one handler for all GPIO interrupts */ - if (pdev->id == 0) - irq_set_chained_handler(port->irq, - mx2_gpio_irq_handler); + /* + * Setup one handler for all GPIO interrupts. Actually setting + * the handler is needed only once, but doing it for every port + * is more robust and easier. + */ + irq_set_chained_handler(port->irq, mx2_gpio_irq_handler); } else { /* setup one handler for each entry */ irq_set_chained_handler(port->irq, mx3_gpio_irq_handler); -- cgit v0.10.2 From 626f9914a371e22f15135f67db6c2aa64adbdacb Mon Sep 17 00:00:00 2001 From: Laxman Dewangan Date: Wed, 4 Jul 2012 20:36:45 +0530 Subject: gpio: tps65910: initialize of_node of gpio_chip Initialize the gpio chip's of_node to the device's node to work with DT based system. Signed-off-by: Laxman Dewangan Signed-off-by: Linus Walleij diff --git a/drivers/gpio/gpio-tps65910.c b/drivers/gpio/gpio-tps65910.c index c1ad288..0749f96 100644 --- a/drivers/gpio/gpio-tps65910.c +++ b/drivers/gpio/gpio-tps65910.c @@ -149,6 +149,7 @@ static int __devinit tps65910_gpio_probe(struct platform_device *pdev) tps65910_gpio->gpio_chip.set = tps65910_gpio_set; tps65910_gpio->gpio_chip.get = tps65910_gpio_get; tps65910_gpio->gpio_chip.dev = &pdev->dev; + tps65910_gpio->gpio_chip.of_node = tps65910->dev->of_node; if (pdata && pdata->gpio_base) tps65910_gpio->gpio_chip.base = pdata->gpio_base; else -- cgit v0.10.2 From 164c33c6adee609b8b9062cce4c10f764d0dce13 Mon Sep 17 00:00:00 2001 From: Salman Qazi Date: Mon, 25 Jun 2012 18:18:15 -0700 Subject: sched: Fix fork() error path to not crash In dup_task_struct(), if arch_dup_task_struct() fails, the clean up code fails to clean up correctly. That's because the clean up code depends on unininitalized ti->task pointer. We fix this by making sure that the task and thread_info know about each other before we attempt to take the error path. Signed-off-by: Salman Qazi Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/20120626011815.11323.5533.stgit@dungbeetle.mtv.corp.google.com Signed-off-by: Ingo Molnar diff --git a/kernel/fork.c b/kernel/fork.c index ab5211b..f00e319 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -304,12 +304,17 @@ static struct task_struct *dup_task_struct(struct task_struct *orig) } err = arch_dup_task_struct(tsk, orig); - if (err) - goto out; + /* + * We defer looking at err, because we will need this setup + * for the clean up path to work correctly. + */ tsk->stack = ti; - setup_thread_stack(tsk, orig); + + if (err) + goto out; + clear_user_return_notifier(tsk); clear_tsk_need_resched(tsk); stackend = end_of_stack(tsk); -- cgit v0.10.2 From 5167e8d5417bf5c322a703d2927daec727ea40dd Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Jun 2012 15:52:09 +0200 Subject: sched/nohz: Rewrite and fix load-avg computation -- again Thanks to Charles Wang for spotting the defects in the current code: - If we go idle during the sample window -- after sampling, we get a negative bias because we can negate our own sample. - If we wake up during the sample window we get a positive bias because we push the sample to a known active period. So rewrite the entire nohz load-avg muck once again, now adding copious documentation to the code. Reported-and-tested-by: Doug Smythies Reported-and-tested-by: Charles Wang Signed-off-by: Peter Zijlstra Cc: Linus Torvalds Cc: Andrew Morton Cc: stable@kernel.org Link: http://lkml.kernel.org/r/1340373782.18025.74.camel@twins [ minor edits ] Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 4059c0f..20cb749 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1909,6 +1909,14 @@ static inline int set_cpus_allowed_ptr(struct task_struct *p, } #endif +#ifdef CONFIG_NO_HZ +void calc_load_enter_idle(void); +void calc_load_exit_idle(void); +#else +static inline void calc_load_enter_idle(void) { } +static inline void calc_load_exit_idle(void) { } +#endif /* CONFIG_NO_HZ */ + #ifndef CONFIG_CPUMASK_OFFSTACK static inline int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask) { diff --git a/kernel/sched/core.c b/kernel/sched/core.c index d5594a4..bb84040 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -2161,11 +2161,73 @@ unsigned long this_cpu_load(void) } +/* + * Global load-average calculations + * + * We take a distributed and async approach to calculating the global load-avg + * in order to minimize overhead. + * + * The global load average is an exponentially decaying average of nr_running + + * nr_uninterruptible. + * + * Once every LOAD_FREQ: + * + * nr_active = 0; + * for_each_possible_cpu(cpu) + * nr_active += cpu_of(cpu)->nr_running + cpu_of(cpu)->nr_uninterruptible; + * + * avenrun[n] = avenrun[0] * exp_n + nr_active * (1 - exp_n) + * + * Due to a number of reasons the above turns in the mess below: + * + * - for_each_possible_cpu() is prohibitively expensive on machines with + * serious number of cpus, therefore we need to take a distributed approach + * to calculating nr_active. + * + * \Sum_i x_i(t) = \Sum_i x_i(t) - x_i(t_0) | x_i(t_0) := 0 + * = \Sum_i { \Sum_j=1 x_i(t_j) - x_i(t_j-1) } + * + * So assuming nr_active := 0 when we start out -- true per definition, we + * can simply take per-cpu deltas and fold those into a global accumulate + * to obtain the same result. See calc_load_fold_active(). + * + * Furthermore, in order to avoid synchronizing all per-cpu delta folding + * across the machine, we assume 10 ticks is sufficient time for every + * cpu to have completed this task. + * + * This places an upper-bound on the IRQ-off latency of the machine. Then + * again, being late doesn't loose the delta, just wrecks the sample. + * + * - cpu_rq()->nr_uninterruptible isn't accurately tracked per-cpu because + * this would add another cross-cpu cacheline miss and atomic operation + * to the wakeup path. Instead we increment on whatever cpu the task ran + * when it went into uninterruptible state and decrement on whatever cpu + * did the wakeup. This means that only the sum of nr_uninterruptible over + * all cpus yields the correct result. + * + * This covers the NO_HZ=n code, for extra head-aches, see the comment below. + */ + /* Variables and functions for calc_load */ static atomic_long_t calc_load_tasks; static unsigned long calc_load_update; unsigned long avenrun[3]; -EXPORT_SYMBOL(avenrun); +EXPORT_SYMBOL(avenrun); /* should be removed */ + +/** + * get_avenrun - get the load average array + * @loads: pointer to dest load array + * @offset: offset to add + * @shift: shift count to shift the result left + * + * These values are estimates at best, so no need for locking. + */ +void get_avenrun(unsigned long *loads, unsigned long offset, int shift) +{ + loads[0] = (avenrun[0] + offset) << shift; + loads[1] = (avenrun[1] + offset) << shift; + loads[2] = (avenrun[2] + offset) << shift; +} static long calc_load_fold_active(struct rq *this_rq) { @@ -2182,6 +2244,9 @@ static long calc_load_fold_active(struct rq *this_rq) return delta; } +/* + * a1 = a0 * e + a * (1 - e) + */ static unsigned long calc_load(unsigned long load, unsigned long exp, unsigned long active) { @@ -2193,30 +2258,118 @@ calc_load(unsigned long load, unsigned long exp, unsigned long active) #ifdef CONFIG_NO_HZ /* - * For NO_HZ we delay the active fold to the next LOAD_FREQ update. + * Handle NO_HZ for the global load-average. + * + * Since the above described distributed algorithm to compute the global + * load-average relies on per-cpu sampling from the tick, it is affected by + * NO_HZ. + * + * The basic idea is to fold the nr_active delta into a global idle-delta upon + * entering NO_HZ state such that we can include this as an 'extra' cpu delta + * when we read the global state. + * + * Obviously reality has to ruin such a delightfully simple scheme: + * + * - When we go NO_HZ idle during the window, we can negate our sample + * contribution, causing under-accounting. + * + * We avoid this by keeping two idle-delta counters and flipping them + * when the window starts, thus separating old and new NO_HZ load. + * + * The only trick is the slight shift in index flip for read vs write. + * + * 0s 5s 10s 15s + * +10 +10 +10 +10 + * |-|-----------|-|-----------|-|-----------|-| + * r:0 0 1 1 0 0 1 1 0 + * w:0 1 1 0 0 1 1 0 0 + * + * This ensures we'll fold the old idle contribution in this window while + * accumlating the new one. + * + * - When we wake up from NO_HZ idle during the window, we push up our + * contribution, since we effectively move our sample point to a known + * busy state. + * + * This is solved by pushing the window forward, and thus skipping the + * sample, for this cpu (effectively using the idle-delta for this cpu which + * was in effect at the time the window opened). This also solves the issue + * of having to deal with a cpu having been in NOHZ idle for multiple + * LOAD_FREQ intervals. * * When making the ILB scale, we should try to pull this in as well. */ -static atomic_long_t calc_load_tasks_idle; +static atomic_long_t calc_load_idle[2]; +static int calc_load_idx; -void calc_load_account_idle(struct rq *this_rq) +static inline int calc_load_write_idx(void) { + int idx = calc_load_idx; + + /* + * See calc_global_nohz(), if we observe the new index, we also + * need to observe the new update time. + */ + smp_rmb(); + + /* + * If the folding window started, make sure we start writing in the + * next idle-delta. + */ + if (!time_before(jiffies, calc_load_update)) + idx++; + + return idx & 1; +} + +static inline int calc_load_read_idx(void) +{ + return calc_load_idx & 1; +} + +void calc_load_enter_idle(void) +{ + struct rq *this_rq = this_rq(); long delta; + /* + * We're going into NOHZ mode, if there's any pending delta, fold it + * into the pending idle delta. + */ delta = calc_load_fold_active(this_rq); - if (delta) - atomic_long_add(delta, &calc_load_tasks_idle); + if (delta) { + int idx = calc_load_write_idx(); + atomic_long_add(delta, &calc_load_idle[idx]); + } } -static long calc_load_fold_idle(void) +void calc_load_exit_idle(void) { - long delta = 0; + struct rq *this_rq = this_rq(); + + /* + * If we're still before the sample window, we're done. + */ + if (time_before(jiffies, this_rq->calc_load_update)) + return; /* - * Its got a race, we don't care... + * We woke inside or after the sample window, this means we're already + * accounted through the nohz accounting, so skip the entire deal and + * sync up for the next window. */ - if (atomic_long_read(&calc_load_tasks_idle)) - delta = atomic_long_xchg(&calc_load_tasks_idle, 0); + this_rq->calc_load_update = calc_load_update; + if (time_before(jiffies, this_rq->calc_load_update + 10)) + this_rq->calc_load_update += LOAD_FREQ; +} + +static long calc_load_fold_idle(void) +{ + int idx = calc_load_read_idx(); + long delta = 0; + + if (atomic_long_read(&calc_load_idle[idx])) + delta = atomic_long_xchg(&calc_load_idle[idx], 0); return delta; } @@ -2302,66 +2455,39 @@ static void calc_global_nohz(void) { long delta, active, n; - /* - * If we crossed a calc_load_update boundary, make sure to fold - * any pending idle changes, the respective CPUs might have - * missed the tick driven calc_load_account_active() update - * due to NO_HZ. - */ - delta = calc_load_fold_idle(); - if (delta) - atomic_long_add(delta, &calc_load_tasks); - - /* - * It could be the one fold was all it took, we done! - */ - if (time_before(jiffies, calc_load_update + 10)) - return; - - /* - * Catch-up, fold however many we are behind still - */ - delta = jiffies - calc_load_update - 10; - n = 1 + (delta / LOAD_FREQ); + if (!time_before(jiffies, calc_load_update + 10)) { + /* + * Catch-up, fold however many we are behind still + */ + delta = jiffies - calc_load_update - 10; + n = 1 + (delta / LOAD_FREQ); - active = atomic_long_read(&calc_load_tasks); - active = active > 0 ? active * FIXED_1 : 0; + active = atomic_long_read(&calc_load_tasks); + active = active > 0 ? active * FIXED_1 : 0; - avenrun[0] = calc_load_n(avenrun[0], EXP_1, active, n); - avenrun[1] = calc_load_n(avenrun[1], EXP_5, active, n); - avenrun[2] = calc_load_n(avenrun[2], EXP_15, active, n); + avenrun[0] = calc_load_n(avenrun[0], EXP_1, active, n); + avenrun[1] = calc_load_n(avenrun[1], EXP_5, active, n); + avenrun[2] = calc_load_n(avenrun[2], EXP_15, active, n); - calc_load_update += n * LOAD_FREQ; -} -#else -void calc_load_account_idle(struct rq *this_rq) -{ -} + calc_load_update += n * LOAD_FREQ; + } -static inline long calc_load_fold_idle(void) -{ - return 0; + /* + * Flip the idle index... + * + * Make sure we first write the new time then flip the index, so that + * calc_load_write_idx() will see the new time when it reads the new + * index, this avoids a double flip messing things up. + */ + smp_wmb(); + calc_load_idx++; } +#else /* !CONFIG_NO_HZ */ -static void calc_global_nohz(void) -{ -} -#endif +static inline long calc_load_fold_idle(void) { return 0; } +static inline void calc_global_nohz(void) { } -/** - * get_avenrun - get the load average array - * @loads: pointer to dest load array - * @offset: offset to add - * @shift: shift count to shift the result left - * - * These values are estimates at best, so no need for locking. - */ -void get_avenrun(unsigned long *loads, unsigned long offset, int shift) -{ - loads[0] = (avenrun[0] + offset) << shift; - loads[1] = (avenrun[1] + offset) << shift; - loads[2] = (avenrun[2] + offset) << shift; -} +#endif /* CONFIG_NO_HZ */ /* * calc_load - update the avenrun load estimates 10 ticks after the @@ -2369,11 +2495,18 @@ void get_avenrun(unsigned long *loads, unsigned long offset, int shift) */ void calc_global_load(unsigned long ticks) { - long active; + long active, delta; if (time_before(jiffies, calc_load_update + 10)) return; + /* + * Fold the 'old' idle-delta to include all NO_HZ cpus. + */ + delta = calc_load_fold_idle(); + if (delta) + atomic_long_add(delta, &calc_load_tasks); + active = atomic_long_read(&calc_load_tasks); active = active > 0 ? active * FIXED_1 : 0; @@ -2384,12 +2517,7 @@ void calc_global_load(unsigned long ticks) calc_load_update += LOAD_FREQ; /* - * Account one period with whatever state we found before - * folding in the nohz state and ageing the entire idle period. - * - * This avoids loosing a sample when we go idle between - * calc_load_account_active() (10 ticks ago) and now and thus - * under-accounting. + * In case we idled for multiple LOAD_FREQ intervals, catch up in bulk. */ calc_global_nohz(); } @@ -2406,7 +2534,6 @@ static void calc_load_account_active(struct rq *this_rq) return; delta = calc_load_fold_active(this_rq); - delta += calc_load_fold_idle(); if (delta) atomic_long_add(delta, &calc_load_tasks); @@ -2414,6 +2541,10 @@ static void calc_load_account_active(struct rq *this_rq) } /* + * End of global load-average stuff + */ + +/* * The exact cpuload at various idx values, calculated at every tick would be * load = (2^idx - 1) / 2^idx * load + 1 / 2^idx * cur_load * diff --git a/kernel/sched/idle_task.c b/kernel/sched/idle_task.c index b44d604..b6baf37 100644 --- a/kernel/sched/idle_task.c +++ b/kernel/sched/idle_task.c @@ -25,7 +25,6 @@ static void check_preempt_curr_idle(struct rq *rq, struct task_struct *p, int fl static struct task_struct *pick_next_task_idle(struct rq *rq) { schedstat_inc(rq, sched_goidle); - calc_load_account_idle(rq); return rq->idle; } diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 6d52cea..55844f2 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -942,8 +942,6 @@ static inline u64 sched_avg_period(void) return (u64)sysctl_sched_time_avg * NSEC_PER_MSEC / 2; } -void calc_load_account_idle(struct rq *this_rq); - #ifdef CONFIG_SCHED_HRTICK /* diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index 8699978..4a08472 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -406,6 +406,7 @@ static void tick_nohz_stop_sched_tick(struct tick_sched *ts) */ if (!ts->tick_stopped) { select_nohz_load_balancer(1); + calc_load_enter_idle(); ts->idle_tick = hrtimer_get_expires(&ts->sched_timer); ts->tick_stopped = 1; @@ -597,6 +598,7 @@ void tick_nohz_idle_exit(void) account_idle_ticks(ticks); #endif + calc_load_exit_idle(); touch_softlockup_watchdog(); /* * Cancel the scheduled timer and restore the tick -- cgit v0.10.2 From 8e16e33c168a6efd0c9f7fa9dd4c1e1db9a74553 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= Date: Mon, 2 Jul 2012 19:53:55 +0200 Subject: USB: option: add ZTE MF60 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Switches into a composite device by ejecting the initial driver CD. The four interfaces are: QCDM, AT, QMI/wwan and mass storage. Let this driver manage the two serial interfaces: T: Bus=02 Lev=01 Prnt=01 Port=01 Cnt=01 Dev#= 28 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1 P: Vendor=19d2 ProdID=1402 Rev= 0.00 S: Manufacturer=ZTE,Incorporated S: Product=ZTE WCDMA Technologies MSM S: SerialNumber=xxxxx C:* #Ifs= 4 Cfg#= 1 Atr=c0 MxPwr=500mA I:* If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=option E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms I:* If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=option E: Ad=82(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms I:* If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan E: Ad=83(I) Atr=03(Int.) MxPS= 64 Ivl=2ms E: Ad=84(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=4ms I:* If#= 3 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage E: Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms Signed-off-by: Bjørn Mork Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index adf8ce7..167aed8 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -554,6 +554,10 @@ static const struct option_blacklist_info net_intf1_blacklist = { .reserved = BIT(1), }; +static const struct option_blacklist_info net_intf2_blacklist = { + .reserved = BIT(2), +}; + static const struct option_blacklist_info net_intf3_blacklist = { .reserved = BIT(3), }; @@ -1099,6 +1103,8 @@ static const struct usb_device_id option_ids[] = { { USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x1298, 0xff, 0xff, 0xff) }, { USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x1299, 0xff, 0xff, 0xff) }, { USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x1300, 0xff, 0xff, 0xff) }, + { USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x1402, 0xff, 0xff, 0xff), + .driver_info = (kernel_ulong_t)&net_intf2_blacklist }, { USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x2002, 0xff, 0xff, 0xff), .driver_info = (kernel_ulong_t)&zte_k3765_z_blacklist }, { USB_DEVICE_AND_INTERFACE_INFO(ZTE_VENDOR_ID, 0x2003, 0xff, 0xff, 0xff) }, -- cgit v0.10.2 From aacef9c561a693341566a6850c451ce3df68cb9a Mon Sep 17 00:00:00 2001 From: Gaosen Zhang Date: Thu, 5 Jul 2012 21:49:00 +0800 Subject: USB: option: Add MEDIATEK product ids Signed-off-by: Gaosen Zhang Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index 167aed8..417ab1b 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -497,6 +497,15 @@ static void option_instat_callback(struct urb *urb); /* MediaTek products */ #define MEDIATEK_VENDOR_ID 0x0e8d +#define MEDIATEK_PRODUCT_DC_1COM 0x00a0 +#define MEDIATEK_PRODUCT_DC_4COM 0x00a5 +#define MEDIATEK_PRODUCT_DC_5COM 0x00a4 +#define MEDIATEK_PRODUCT_7208_1COM 0x7101 +#define MEDIATEK_PRODUCT_7208_2COM 0x7102 +#define MEDIATEK_PRODUCT_FP_1COM 0x0003 +#define MEDIATEK_PRODUCT_FP_2COM 0x0023 +#define MEDIATEK_PRODUCT_FPDC_1COM 0x0043 +#define MEDIATEK_PRODUCT_FPDC_2COM 0x0033 /* Cellient products */ #define CELLIENT_VENDOR_ID 0x2692 @@ -1246,6 +1255,17 @@ static const struct usb_device_id option_ids[] = { { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, 0x00a1, 0xff, 0x02, 0x01) }, { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, 0x00a2, 0xff, 0x00, 0x00) }, { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, 0x00a2, 0xff, 0x02, 0x01) }, /* MediaTek MT6276M modem & app port */ + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_DC_1COM, 0x0a, 0x00, 0x00) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_DC_5COM, 0xff, 0x02, 0x01) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_DC_5COM, 0xff, 0x00, 0x00) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_DC_4COM, 0xff, 0x02, 0x01) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_DC_4COM, 0xff, 0x00, 0x00) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_7208_1COM, 0x02, 0x00, 0x00) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_7208_2COM, 0x02, 0x02, 0x01) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_FP_1COM, 0x0a, 0x00, 0x00) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_FP_2COM, 0x0a, 0x00, 0x00) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_FPDC_1COM, 0x0a, 0x00, 0x00) }, + { USB_DEVICE_AND_INTERFACE_INFO(MEDIATEK_VENDOR_ID, MEDIATEK_PRODUCT_FPDC_2COM, 0x0a, 0x00, 0x00) }, { USB_DEVICE(CELLIENT_VENDOR_ID, CELLIENT_PRODUCT_MEN200) }, { } /* Terminating entry */ }; -- cgit v0.10.2 From b7d28e32c93801d60c1a7a817f774a02b7bdde43 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Mon, 2 Jul 2012 12:34:24 +0200 Subject: USB: metro-usb: fix tty_flip_buffer_push use Do not set low_latency flag at open as tty_flip_buffer_push must not be called in IRQ context with low_latency set. Cc: stable@vger.kernel.org Signed-off-by: Johan Hovold Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/metro-usb.c b/drivers/usb/serial/metro-usb.c index 81423f7..d47eb06 100644 --- a/drivers/usb/serial/metro-usb.c +++ b/drivers/usb/serial/metro-usb.c @@ -222,14 +222,6 @@ static int metrousb_open(struct tty_struct *tty, struct usb_serial_port *port) metro_priv->throttled = 0; spin_unlock_irqrestore(&metro_priv->lock, flags); - /* - * Force low_latency on so that our tty_push actually forces the data - * through, otherwise it is scheduled, and with high data rates (like - * with OHCI) data can get lost. - */ - if (tty) - tty->low_latency = 1; - /* Clear the urb pipe. */ usb_clear_halt(serial->dev, port->interrupt_in_urb->pipe); -- cgit v0.10.2 From b086b6b10d9f182cd8d2f0dcfd7fd11edba93fc9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= Date: Mon, 2 Jul 2012 10:33:14 +0200 Subject: USB: cdc-wdm: fix lockup on error in wdm_read MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Clear the WDM_READ flag on empty reads to avoid running forever in an infinite tight loop, causing lockups: Jul 1 21:58:11 nemi kernel: [ 3658.898647] qmi_wwan 2-1:1.2: Unexpected error -71 Jul 1 21:58:36 nemi kernel: [ 3684.072021] BUG: soft lockup - CPU#0 stuck for 23s! [qmi.pl:12235] Jul 1 21:58:36 nemi kernel: [ 3684.072212] CPU 0 Jul 1 21:58:36 nemi kernel: [ 3684.072355] Jul 1 21:58:36 nemi kernel: [ 3684.072367] Pid: 12235, comm: qmi.pl Tainted: P O 3.5.0-rc2+ #13 LENOVO 2776LEG/2776LEG Jul 1 21:58:36 nemi kernel: [ 3684.072383] RIP: 0010:[] [] spin_unlock_irq+0x8/0xc [cdc_wdm] Jul 1 21:58:36 nemi kernel: [ 3684.072388] RSP: 0018:ffff88022dca1e70 EFLAGS: 00000282 Jul 1 21:58:36 nemi kernel: [ 3684.072393] RAX: ffff88022fc3f650 RBX: ffffffff811c56f7 RCX: 00000001000ce8c1 Jul 1 21:58:36 nemi kernel: [ 3684.072398] RDX: 0000000000000010 RSI: 000000000267d810 RDI: ffff88022fc3f650 Jul 1 21:58:36 nemi kernel: [ 3684.072403] RBP: ffff88022dca1eb0 R08: ffffffffa063578e R09: 0000000000000000 Jul 1 21:58:36 nemi kernel: [ 3684.072407] R10: 0000000000000008 R11: 0000000000000246 R12: 0000000000000002 Jul 1 21:58:36 nemi kernel: [ 3684.072412] R13: 0000000000000246 R14: ffffffff00000002 R15: ffff8802281d8c88 Jul 1 21:58:36 nemi kernel: [ 3684.072418] FS: 00007f666a260700(0000) GS:ffff88023bc00000(0000) knlGS:0000000000000000 Jul 1 21:58:36 nemi kernel: [ 3684.072423] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 Jul 1 21:58:36 nemi kernel: [ 3684.072428] CR2: 000000000270d9d8 CR3: 000000022e865000 CR4: 00000000000007f0 Jul 1 21:58:36 nemi kernel: [ 3684.072433] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 Jul 1 21:58:36 nemi kernel: [ 3684.072438] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Jul 1 21:58:36 nemi kernel: [ 3684.072444] Process qmi.pl (pid: 12235, threadinfo ffff88022dca0000, task ffff88022ff76380) Jul 1 21:58:36 nemi kernel: [ 3684.072448] Stack: Jul 1 21:58:36 nemi kernel: [ 3684.072458] ffffffffa063592e 0000000100020000 ffff88022fc3f650 ffff88022fc3f6a8 Jul 1 21:58:36 nemi kernel: [ 3684.072466] 0000000000000200 0000000100000000 000000000267d810 0000000000000000 Jul 1 21:58:36 nemi kernel: [ 3684.072475] 0000000000000000 ffff880212cfb6d0 0000000000000200 ffff880212cfb6c0 Jul 1 21:58:36 nemi kernel: [ 3684.072479] Call Trace: Jul 1 21:58:36 nemi kernel: [ 3684.072489] [] ? wdm_read+0x1a0/0x263 [cdc_wdm] Jul 1 21:58:36 nemi kernel: [ 3684.072500] [] ? vfs_read+0xa1/0xfb Jul 1 21:58:36 nemi kernel: [ 3684.072509] [] ? alarm_setitimer+0x35/0x64 Jul 1 21:58:36 nemi kernel: [ 3684.072517] [] ? sys_read+0x45/0x6e Jul 1 21:58:36 nemi kernel: [ 3684.072525] [] ? system_call_fastpath+0x16/0x1b Jul 1 21:58:36 nemi kernel: [ 3684.072557] Code: <66> 66 90 c3 83 ff ed 89 f8 74 16 7f 06 83 ff a1 75 0a c3 83 ff f4 The WDM_READ flag is normally cleared by wdm_int_callback before resubmitting the read urb, and set by wdm_in_callback when this urb returns with data or an error. But a crashing device may cause both a read error and cancelling all urbs. Make sure that the flag is cleared by wdm_read if the buffer is empty. We don't clear the flag on errors, as there may be pending data in the buffer which should be processed. The flag will instead be cleared on the next wdm_read call. Cc: stable Signed-off-by: Bjørn Mork Acked-by: Oliver Neukum Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/class/cdc-wdm.c b/drivers/usb/class/cdc-wdm.c index 8fd398d..ee46927 100644 --- a/drivers/usb/class/cdc-wdm.c +++ b/drivers/usb/class/cdc-wdm.c @@ -500,6 +500,8 @@ retry: goto retry; } if (!desc->reslength) { /* zero length read */ + dev_dbg(&desc->intf->dev, "%s: zero length - clearing WDM_READ\n", __func__); + clear_bit(WDM_READ, &desc->flags); spin_unlock_irq(&desc->iuspin); goto retry; } -- cgit v0.10.2 From 006c7f18449a06027b0165e938c67b3a029813c9 Mon Sep 17 00:00:00 2001 From: Paul Walmsley Date: Wed, 4 Jul 2012 05:22:53 -0600 Subject: ARM: OMAP2+: hwmod code/clockdomain data: fix 32K sync timer MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Kevin discovered that commit c8d82ff68fb6873691536cf33021977efbf5593c ("ARM: OMAP2/3: hwmod data: Add 32k-sync timer data to hwmod database") broke CORE idle on OMAP3. This prevents device low power states. The root cause is that the 32K sync timer IP block does not support smart-idle mode[1], and so the hwmod code keeps the IP block in no-idle mode while it is active. This in turn prevents the WKUP clockdomain from transitioning to idle. There is a hardcoded sleep dependency that prevents the CORE_L3 and CORE_CM clockdomains from transitioning to idle when the WKUP clockdomain is active[2], so the chip cannot enter any device low power states. It turns out that there is no need to take the 32k sync timer out of idle. The IP block itself probably does not have any native idle handling at all, due to its simplicity. Furthermore, the PRCM will never request target idle for this IP block while the kernel is running, due to the sleep dependency that prevents the WKUP clockdomain from idling while the CORE_L3 clockdomain is active. So we can safely leave the 32k sync timer in target-force-idle mode, even while we continue to access it. This workaround is implemented by defining a new clockdomain flag, CLKDM_ACTIVE_WITH_MPU, that indicates that the clockdomain is guaranteed to be active whenever the MPU is inactive. If an IP block's main functional clock exists inside this clockdomain, and the IP block does not support smart-idle modes, then the hwmod code will place the IP block into target force-idle mode even when enabled. The WKUP clockdomains on OMAP3/4 are marked with this flag. (On OMAP2xxx, no OCP header existed on the 32k sync timer.) Other clockdomains also should be marked with this flag, but those changes are deferred until a later merge window, to create a minimal fix. Another theoretically clean fix for this problem would be to implement PM runtime-based control for 32k sync timer accesses. These PM runtime calls would need to located in a custom clocksource, since the 32k sync timer is currently used as an MMIO clocksource. But in practice, there would be little benefit to doing so; and there would be some cost, due to the addition of unnecessary lines of code and the additional CPU overhead of the PM runtime and hwmod code - unnecessary in this case. Another possible fix would have been to modify the pm34xx.c code to force the IP block idle before entering WFI. But this would not have been an acceptable approach: we are trying to remove this type of centralized IP block idle control from the PM code. This patch is a collaboration between Kevin Hilman and Paul Walmsley . Thanks to Vaibhav Hiremath for providing comments on an earlier version of this patch. Thanks to Tero Kristo for identifying a bug in an earlier version of this patch. Thanks to Benoît Cousson for identifying some bugs in several versions of this patch and for implementation comments. References: 1. Table 16-96 "REG_32KSYNCNT_SYSCONFIG" of the OMAP34xx TRM Rev. ZU (SWPU223U), available from: http://www.ti.com/pdfs/wtbu/OMAP34x_ES3.1.x_PUBLIC_TRM_vzU.zip 2. Table 4-72 "Sleep Dependencies" of the OMAP34xx TRM Rev. ZU (SWPU223U) 3. ibid. Cc: Tony Lindgren Cc: Vaibhav Hiremath Cc: Benoît Cousson Cc: Tero Kristo Tested-by: Kevin Hilman Signed-off-by: Paul Walmsley Signed-off-by: Kevin Hilman diff --git a/arch/arm/mach-omap2/clockdomain.h b/arch/arm/mach-omap2/clockdomain.h index f7b58609..6227e95 100644 --- a/arch/arm/mach-omap2/clockdomain.h +++ b/arch/arm/mach-omap2/clockdomain.h @@ -31,12 +31,16 @@ * * CLKDM_NO_AUTODEPS: Prevent "autodeps" from being added/removed from this * clockdomain. (Currently, this applies to OMAP3 clockdomains only.) + * CLKDM_ACTIVE_WITH_MPU: The PRCM guarantees that this clockdomain is + * active whenever the MPU is active. True for interconnects and + * the WKUP clockdomains. */ #define CLKDM_CAN_FORCE_SLEEP (1 << 0) #define CLKDM_CAN_FORCE_WAKEUP (1 << 1) #define CLKDM_CAN_ENABLE_AUTO (1 << 2) #define CLKDM_CAN_DISABLE_AUTO (1 << 3) #define CLKDM_NO_AUTODEPS (1 << 4) +#define CLKDM_ACTIVE_WITH_MPU (1 << 5) #define CLKDM_CAN_HWSUP (CLKDM_CAN_ENABLE_AUTO | CLKDM_CAN_DISABLE_AUTO) #define CLKDM_CAN_SWSUP (CLKDM_CAN_FORCE_SLEEP | CLKDM_CAN_FORCE_WAKEUP) diff --git a/arch/arm/mach-omap2/clockdomains2xxx_3xxx_data.c b/arch/arm/mach-omap2/clockdomains2xxx_3xxx_data.c index 839145e1..4972219 100644 --- a/arch/arm/mach-omap2/clockdomains2xxx_3xxx_data.c +++ b/arch/arm/mach-omap2/clockdomains2xxx_3xxx_data.c @@ -88,4 +88,5 @@ struct clockdomain wkup_common_clkdm = { .name = "wkup_clkdm", .pwrdm = { .name = "wkup_pwrdm" }, .dep_bit = OMAP_EN_WKUP_SHIFT, + .flags = CLKDM_ACTIVE_WITH_MPU, }; diff --git a/arch/arm/mach-omap2/clockdomains44xx_data.c b/arch/arm/mach-omap2/clockdomains44xx_data.c index c534258..7f2133a 100644 --- a/arch/arm/mach-omap2/clockdomains44xx_data.c +++ b/arch/arm/mach-omap2/clockdomains44xx_data.c @@ -381,7 +381,7 @@ static struct clockdomain l4_wkup_44xx_clkdm = { .cm_inst = OMAP4430_PRM_WKUP_CM_INST, .clkdm_offs = OMAP4430_PRM_WKUP_CM_WKUP_CDOFFS, .dep_bit = OMAP4430_L4WKUP_STATDEP_SHIFT, - .flags = CLKDM_CAN_HWSUP, + .flags = CLKDM_CAN_HWSUP | CLKDM_ACTIVE_WITH_MPU, }; static struct clockdomain emu_sys_44xx_clkdm = { diff --git a/arch/arm/mach-omap2/omap_hwmod.c b/arch/arm/mach-omap2/omap_hwmod.c index 7731936..2d710f5 100644 --- a/arch/arm/mach-omap2/omap_hwmod.c +++ b/arch/arm/mach-omap2/omap_hwmod.c @@ -1124,15 +1124,18 @@ static struct omap_hwmod_addr_space * __init _find_mpu_rt_addr_space(struct omap * _enable_sysc - try to bring a module out of idle via OCP_SYSCONFIG * @oh: struct omap_hwmod * * - * If module is marked as SWSUP_SIDLE, force the module out of slave - * idle; otherwise, configure it for smart-idle. If module is marked - * as SWSUP_MSUSPEND, force the module out of master standby; - * otherwise, configure it for smart-standby. No return value. + * Ensure that the OCP_SYSCONFIG register for the IP block represented + * by @oh is set to indicate to the PRCM that the IP block is active. + * Usually this means placing the module into smart-idle mode and + * smart-standby, but if there is a bug in the automatic idle handling + * for the IP block, it may need to be placed into the force-idle or + * no-idle variants of these modes. No return value. */ static void _enable_sysc(struct omap_hwmod *oh) { u8 idlemode, sf; u32 v; + bool clkdm_act; if (!oh->class->sysc) return; @@ -1141,8 +1144,16 @@ static void _enable_sysc(struct omap_hwmod *oh) sf = oh->class->sysc->sysc_flags; if (sf & SYSC_HAS_SIDLEMODE) { - idlemode = (oh->flags & HWMOD_SWSUP_SIDLE) ? - HWMOD_IDLEMODE_NO : HWMOD_IDLEMODE_SMART; + clkdm_act = ((oh->clkdm && + oh->clkdm->flags & CLKDM_ACTIVE_WITH_MPU) || + (oh->_clk && oh->_clk->clkdm && + oh->_clk->clkdm->flags & CLKDM_ACTIVE_WITH_MPU)); + if (clkdm_act && !(oh->class->sysc->idlemodes & + (SIDLE_SMART | SIDLE_SMART_WKUP))) + idlemode = HWMOD_IDLEMODE_FORCE; + else + idlemode = (oh->flags & HWMOD_SWSUP_SIDLE) ? + HWMOD_IDLEMODE_NO : HWMOD_IDLEMODE_SMART; _set_slave_idlemode(oh, idlemode, &v); } @@ -1208,8 +1219,13 @@ static void _idle_sysc(struct omap_hwmod *oh) sf = oh->class->sysc->sysc_flags; if (sf & SYSC_HAS_SIDLEMODE) { - idlemode = (oh->flags & HWMOD_SWSUP_SIDLE) ? - HWMOD_IDLEMODE_FORCE : HWMOD_IDLEMODE_SMART; + /* XXX What about HWMOD_IDLEMODE_SMART_WKUP? */ + if (oh->flags & HWMOD_SWSUP_SIDLE || + !(oh->class->sysc->idlemodes & + (SIDLE_SMART | SIDLE_SMART_WKUP))) + idlemode = HWMOD_IDLEMODE_FORCE; + else + idlemode = HWMOD_IDLEMODE_SMART; _set_slave_idlemode(oh, idlemode, &v); } -- cgit v0.10.2 From 9e9b594661e8dee54955607277c937c3ff3a5f01 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 6 Jul 2012 08:11:43 +0200 Subject: ALSA: usb-audio: Fix the first PCM interface assignment In the new PCM streaming logic, the interface number is assigned to usb stream instance (subs->interface) after the format and rate setups are succeeded, but some codes are still passing subs->interface as the reference to helper functions. This leads to initializing with an invalid iface number (-1). This patch replaces the wrong references with the ones from the target fmt correctly. Signed-off-by: Takashi Iwai diff --git a/sound/usb/pcm.c b/sound/usb/pcm.c index 54607f8..f0ede13 100644 --- a/sound/usb/pcm.c +++ b/sound/usb/pcm.c @@ -387,7 +387,7 @@ add_sync_ep: subs->data_endpoint->sync_master = subs->sync_endpoint; } - if ((err = snd_usb_init_pitch(subs->stream->chip, subs->interface, alts, fmt)) < 0) + if ((err = snd_usb_init_pitch(subs->stream->chip, fmt->iface, alts, fmt)) < 0) return err; subs->cur_audiofmt = fmt; @@ -450,7 +450,7 @@ static int snd_usb_hw_params(struct snd_pcm_substream *substream, struct usb_interface *iface; iface = usb_ifnum_to_if(subs->dev, fmt->iface); alts = &iface->altsetting[fmt->altset_idx]; - ret = snd_usb_init_sample_rate(subs->stream->chip, subs->interface, alts, fmt, rate); + ret = snd_usb_init_sample_rate(subs->stream->chip, fmt->iface, alts, fmt, rate); if (ret < 0) return ret; subs->cur_rate = rate; -- cgit v0.10.2 From 6a6dccba2fdc2a69f1f36b8f1c0acc8598e7221b Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Thu, 5 Jul 2012 15:52:23 +0530 Subject: mm: cma: don't replace lowmem pages with highmem The filesystem layer expects pages in the block device's mapping to not be in highmem (the mapping's gfp mask is set in bdget()), but CMA can currently replace lowmem pages with highmem pages, leading to crashes in filesystem code such as the one below: Unable to handle kernel NULL pointer dereference at virtual address 00000400 pgd = c0c98000 [00000400] *pgd=00c91831, *pte=00000000, *ppte=00000000 Internal error: Oops: 817 [#1] PREEMPT SMP ARM CPU: 0 Not tainted (3.5.0-rc5+ #80) PC is at __memzero+0x24/0x80 ... Process fsstress (pid: 323, stack limit = 0xc0cbc2f0) Backtrace: [] (ext4_getblk+0x0/0x180) from [] (ext4_bread+0x1c/0x98) [] (ext4_bread+0x0/0x98) from [] (ext4_mkdir+0x160/0x3bc) r4:c15337f0 [] (ext4_mkdir+0x0/0x3bc) from [] (vfs_mkdir+0x8c/0x98) [] (vfs_mkdir+0x0/0x98) from [] (sys_mkdirat+0x74/0xac) r6:00000000 r5:c152eb40 r4:000001ff r3:c14b43f0 [] (sys_mkdirat+0x0/0xac) from [] (sys_mkdir+0x20/0x24) r6:beccdcf0 r5:00074000 r4:beccdbbc [] (sys_mkdir+0x0/0x24) from [] (ret_fast_syscall+0x0/0x30) Fix this by replacing only highmem pages with highmem. Reported-by: Laura Abbott Signed-off-by: Rabin Vincent Acked-by: Michal Nazarewicz Signed-off-by: Marek Szyprowski diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 4403009..4a4f921 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -5635,7 +5635,12 @@ static struct page * __alloc_contig_migrate_alloc(struct page *page, unsigned long private, int **resultp) { - return alloc_page(GFP_HIGHUSER_MOVABLE); + gfp_t gfp_mask = GFP_USER | __GFP_MOVABLE; + + if (PageHighMem(page)) + gfp_mask |= __GFP_HIGHMEM; + + return alloc_page(gfp_mask); } /* [start, end) must belong to a single zone. */ -- cgit v0.10.2 From cc2caea5b6152b8ce66dc2bbe83dc72b60612da8 Mon Sep 17 00:00:00 2001 From: Marek Szyprowski Date: Fri, 6 Jul 2012 12:02:04 +0200 Subject: mm: cma: fix condition check when setting global cma area dev_set_cma_area incorrectly assigned cma to global area on first call due to incorrect check. This patch fixes this issue. Signed-off-by: Marek Szyprowski diff --git a/include/asm-generic/dma-contiguous.h b/include/asm-generic/dma-contiguous.h index c544356..294b1e7 100644 --- a/include/asm-generic/dma-contiguous.h +++ b/include/asm-generic/dma-contiguous.h @@ -18,7 +18,7 @@ static inline void dev_set_cma_area(struct device *dev, struct cma *cma) { if (dev) dev->cma_area = cma; - if (!dev || !dma_contiguous_default_area) + if (!dev && !dma_contiguous_default_area) dma_contiguous_default_area = cma; } -- cgit v0.10.2 From 95c0d71dcb853c2eca5f2231ebbd4c1d3af775b7 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Tue, 3 Jul 2012 23:37:31 +0900 Subject: MAINTAINERS/sched: Update scheduler file pattern The commit 391e43da797a ("sched: Move all scheduler bits into kernel/sched/") moved all scheduler codes to the kernel/sched/ directory, but missed the MAINTAINERS. Since it still expects files from kernel/ directory, get_maintainer script has to rely on the git (log) fallback mechanism. $ scripts/get_maintainer.pl -f kernel/sched/core.c --nogit-fallback linux-kernel@vger.kernel.org (open list) With this patch: $ scripts/get_maintainer.pl -f kernel/sched/core.c --nogit-fallback Ingo Molnar (maintainer:SCHEDULER) Peter Zijlstra (maintainer:SCHEDULER) linux-kernel@vger.kernel.org (open list) Signed-off-by: Namhyung Kim Acked-by: Peter Zijlstra Link: http://lkml.kernel.org/r/1341326251-4140-1-git-send-email-namhyung@kernel.org Signed-off-by: Ingo Molnar diff --git a/MAINTAINERS b/MAINTAINERS index 03df1d1..214d754 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5909,7 +5909,7 @@ M: Ingo Molnar M: Peter Zijlstra T: git git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git sched/core S: Maintained -F: kernel/sched* +F: kernel/sched/ F: include/linux/sched.h SCORE ARCHITECTURE -- cgit v0.10.2 From 48f8b641297df49021093763a3271119a84990a2 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 9 Jun 2012 19:08:25 +0300 Subject: mtd: cafe_nand: fix an & vs | mistake The intent here was clearly to set result to true if the 0x40000000 flag was set. But instead there was a | vs & typo and we always set result to true. Artem: check the spec at wiki.laptop.org/images/5/5c/88ALP01_Datasheet_July_2007.pdf and this fix looks correct. Signed-off-by: Dan Carpenter Cc: stable@vger.kernel.org Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/cafe_nand.c b/drivers/mtd/nand/cafe_nand.c index 41371ba..f3f6cfe 100644 --- a/drivers/mtd/nand/cafe_nand.c +++ b/drivers/mtd/nand/cafe_nand.c @@ -102,7 +102,7 @@ static const char *part_probes[] = { "cmdlinepart", "RedBoot", NULL }; static int cafe_device_ready(struct mtd_info *mtd) { struct cafe_priv *cafe = mtd->priv; - int result = !!(cafe_readl(cafe, NAND_STATUS) | 0x40000000); + int result = !!(cafe_readl(cafe, NAND_STATUS) & 0x40000000); uint32_t irqs = cafe_readl(cafe, NAND_IRQ); cafe_writel(cafe, irqs, NAND_IRQ); -- cgit v0.10.2 From 096bcc231fd263bc8df215f0d616b08e3696c6db Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Tue, 29 May 2012 10:16:09 +0200 Subject: mtd: mxc_nand: use 32bit copy functions MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The following commit changes the function used to copy from/to the hardware buffer to memcpy_[from|to]io. This does not work since the hardware cannot handle the byte accesses used by these functions. Instead of reverting this patch introduce 32bit correspondents of these functions. | commit 5775ba36ea9c760c2d7e697dac04f2f7fc95aa62 | Author: Uwe Kleine-König | Date: Tue Apr 24 10:05:22 2012 +0200 | | mtd: mxc_nand: fix several sparse warnings about incorrect address space | | Signed-off-by: Uwe Kleine-König | Signed-off-by: Artem Bityutskiy Signed-off-by: Sascha Hauer Cc: Uwe Kleine-König Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/mxc_nand.c b/drivers/mtd/nand/mxc_nand.c index c58e6a9..6acc790 100644 --- a/drivers/mtd/nand/mxc_nand.c +++ b/drivers/mtd/nand/mxc_nand.c @@ -273,6 +273,26 @@ static struct nand_ecclayout nandv2_hw_eccoob_4k = { static const char *part_probes[] = { "RedBoot", "cmdlinepart", "ofpart", NULL }; +static void memcpy32_fromio(void *trg, const void __iomem *src, size_t size) +{ + int i; + u32 *t = trg; + const __iomem u32 *s = src; + + for (i = 0; i < (size >> 2); i++) + *t++ = __raw_readl(s++); +} + +static void memcpy32_toio(void __iomem *trg, const void *src, int size) +{ + int i; + u32 __iomem *t = trg; + const u32 *s = src; + + for (i = 0; i < (size >> 2); i++) + __raw_writel(*s++, t++); +} + static int check_int_v3(struct mxc_nand_host *host) { uint32_t tmp; @@ -519,7 +539,7 @@ static void send_read_id_v3(struct mxc_nand_host *host) wait_op_done(host, true); - memcpy_fromio(host->data_buf, host->main_area0, 16); + memcpy32_fromio(host->data_buf, host->main_area0, 16); } /* Request the NANDFC to perform a read of the NAND device ID. */ @@ -535,7 +555,7 @@ static void send_read_id_v1_v2(struct mxc_nand_host *host) /* Wait for operation to complete */ wait_op_done(host, true); - memcpy_fromio(host->data_buf, host->main_area0, 16); + memcpy32_fromio(host->data_buf, host->main_area0, 16); if (this->options & NAND_BUSWIDTH_16) { /* compress the ID info */ @@ -797,16 +817,16 @@ static void copy_spare(struct mtd_info *mtd, bool bfrom) if (bfrom) { for (i = 0; i < n - 1; i++) - memcpy_fromio(d + i * j, s + i * t, j); + memcpy32_fromio(d + i * j, s + i * t, j); /* the last section */ - memcpy_fromio(d + i * j, s + i * t, mtd->oobsize - i * j); + memcpy32_fromio(d + i * j, s + i * t, mtd->oobsize - i * j); } else { for (i = 0; i < n - 1; i++) - memcpy_toio(&s[i * t], &d[i * j], j); + memcpy32_toio(&s[i * t], &d[i * j], j); /* the last section */ - memcpy_toio(&s[i * t], &d[i * j], mtd->oobsize - i * j); + memcpy32_toio(&s[i * t], &d[i * j], mtd->oobsize - i * j); } } @@ -1070,7 +1090,8 @@ static void mxc_nand_command(struct mtd_info *mtd, unsigned command, host->devtype_data->send_page(mtd, NFC_OUTPUT); - memcpy_fromio(host->data_buf, host->main_area0, mtd->writesize); + memcpy32_fromio(host->data_buf, host->main_area0, + mtd->writesize); copy_spare(mtd, true); break; @@ -1086,7 +1107,7 @@ static void mxc_nand_command(struct mtd_info *mtd, unsigned command, break; case NAND_CMD_PAGEPROG: - memcpy_toio(host->main_area0, host->data_buf, mtd->writesize); + memcpy32_toio(host->main_area0, host->data_buf, mtd->writesize); copy_spare(mtd, false); host->devtype_data->send_page(mtd, NFC_INPUT); host->devtype_data->send_cmd(host, command, true); -- cgit v0.10.2 From 6023813a2d5949ba368e7df464f2ccb649719777 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Tue, 26 Jun 2012 17:26:16 +0200 Subject: mtd: gpmi-nand: fix read page when reading to vmalloced area The gpmi-nand driver uses virt_addr_valid() to check whether a buffer is suitable for dma. If it's not, a driver allocated buffer is used instead. Then after a page read the driver allocated buffer must be copied to the user supplied buffer. This does not happen since commit 7725cc85932bd02dd12c23108e0ef748c551ccba. This patch fixes the issue. The bug is encountered with UBI which uses a vmalloced buffer for the volume table. Signed-off-by: Sascha Hauer Tested-by: snijsure@grid-net.com Acked-by: Huang Shijie Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c index a05b7b4..a6cad5c 100644 --- a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c +++ b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c @@ -920,12 +920,12 @@ static int gpmi_ecc_read_page(struct mtd_info *mtd, struct nand_chip *chip, */ memset(chip->oob_poi, ~0, mtd->oobsize); chip->oob_poi[0] = ((uint8_t *) auxiliary_virt)[0]; - - read_page_swap_end(this, buf, mtd->writesize, - this->payload_virt, this->payload_phys, - nfc_geo->payload_size, - payload_virt, payload_phys); } + + read_page_swap_end(this, buf, mtd->writesize, + this->payload_virt, this->payload_phys, + nfc_geo->payload_size, + payload_virt, payload_phys); exit_nfc: return ret; } -- cgit v0.10.2 From 021796b892dcad45743f3196e3eef7222870dd55 Mon Sep 17 00:00:00 2001 From: Mike Dunn Date: Tue, 22 May 2012 11:03:42 -0700 Subject: mtd: ABI documentation: clarification of bitflip_threshold The -EUCLEAN return value applies to mtd_read_oob() as well as mtd_read(), but only mtd_read() was mentioned in the blurd on bitflip_threshold in the ABI documentation. Signed-off-by: Mike Dunn Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/Documentation/ABI/testing/sysfs-class-mtd b/Documentation/ABI/testing/sysfs-class-mtd index db1ad7e..938ef71 100644 --- a/Documentation/ABI/testing/sysfs-class-mtd +++ b/Documentation/ABI/testing/sysfs-class-mtd @@ -142,13 +142,14 @@ KernelVersion: 3.4 Contact: linux-mtd@lists.infradead.org Description: This allows the user to examine and adjust the criteria by which - mtd returns -EUCLEAN from mtd_read(). If the maximum number of - bit errors that were corrected on any single region comprising - an ecc step (as reported by the driver) equals or exceeds this - value, -EUCLEAN is returned. Otherwise, absent an error, 0 is - returned. Higher layers (e.g., UBI) use this return code as an - indication that an erase block may be degrading and should be - scrutinized as a candidate for being marked as bad. + mtd returns -EUCLEAN from mtd_read() and mtd_read_oob(). If the + maximum number of bit errors that were corrected on any single + region comprising an ecc step (as reported by the driver) equals + or exceeds this value, -EUCLEAN is returned. Otherwise, absent + an error, 0 is returned. Higher layers (e.g., UBI) use this + return code as an indication that an erase block may be + degrading and should be scrutinized as a candidate for being + marked as bad. The initial value may be specified by the flash device driver. If not, then the default value is ecc_strength. @@ -167,7 +168,7 @@ Description: block degradation, but high enough to avoid the consequences of a persistent return value of -EUCLEAN on devices where sticky bitflips occur. Note that if bitflip_threshold exceeds - ecc_strength, -EUCLEAN is never returned by mtd_read(). + ecc_strength, -EUCLEAN is never returned by the read operations. Conversely, if bitflip_threshold is zero, -EUCLEAN is always returned, absent a hard error. -- cgit v0.10.2 From 596fd46268634082314b3af1ded4612e1b7f3f03 Mon Sep 17 00:00:00 2001 From: Herton Ronaldo Krzesinski Date: Wed, 16 May 2012 16:21:52 -0300 Subject: mtd: nandsim: don't open code a do_div helper We don't need to open code the divide function, just use div_u64 that already exists and do the same job. While this is a straightforward clean up, there is more to that, the real motivation for this. While building on a cross compiling environment in armel, using gcc 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5), I was getting the following build error: ERROR: "__aeabi_uldivmod" [drivers/mtd/nand/nandsim.ko] undefined! After investigating with objdump and hand built assembly version generated with the compiler, I narrowed __aeabi_uldivmod as being generated from the divide function. When nandsim.c is built with -fno-inline-functions-called-once, that happens when CONFIG_DEBUG_SECTION_MISMATCH is enabled, the do_div optimization in arch/arm/include/asm/div64.h doesn't work as expected with the open coded divide function: even if the do_div we are using doesn't have a constant divisor, the compiler still includes the else parts of the optimized do_div macro, and translates the divisions there to use __aeabi_uldivmod, instead of only calling __do_div_asm -> __do_div64 and optimizing/removing everything else out. So to reproduce, gcc 4.6 plus CONFIG_DEBUG_SECTION_MISMATCH=y and CONFIG_MTD_NAND_NANDSIM=m should do it, building on armel. After this change, the compiler does the intended thing even with -fno-inline-functions-called-once, and optimizes out as expected the constant handling in the optimized do_div on arm. As this also avoids a build issue, I'm marking for Stable, as I think is applicable for this case. Signed-off-by: Herton Ronaldo Krzesinski Cc: stable@vger.kernel.org Acked-by: Nicolas Pitre Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse diff --git a/drivers/mtd/nand/nandsim.c b/drivers/mtd/nand/nandsim.c index 6cc8fbf..cf0cd31 100644 --- a/drivers/mtd/nand/nandsim.c +++ b/drivers/mtd/nand/nandsim.c @@ -28,7 +28,7 @@ #include #include #include -#include +#include #include #include #include @@ -546,12 +546,6 @@ static char *get_partition_name(int i) return kstrdup(buf, GFP_KERNEL); } -static uint64_t divide(uint64_t n, uint32_t d) -{ - do_div(n, d); - return n; -} - /* * Initialize the nandsim structure. * @@ -580,7 +574,7 @@ static int init_nandsim(struct mtd_info *mtd) ns->geom.oobsz = mtd->oobsize; ns->geom.secsz = mtd->erasesize; ns->geom.pgszoob = ns->geom.pgsz + ns->geom.oobsz; - ns->geom.pgnum = divide(ns->geom.totsz, ns->geom.pgsz); + ns->geom.pgnum = div_u64(ns->geom.totsz, ns->geom.pgsz); ns->geom.totszoob = ns->geom.totsz + (uint64_t)ns->geom.pgnum * ns->geom.oobsz; ns->geom.secshift = ffs(ns->geom.secsz) - 1; ns->geom.pgshift = chip->page_shift; @@ -921,7 +915,7 @@ static int setup_wear_reporting(struct mtd_info *mtd) if (!rptwear) return 0; - wear_eb_count = divide(mtd->size, mtd->erasesize); + wear_eb_count = div_u64(mtd->size, mtd->erasesize); mem = wear_eb_count * sizeof(unsigned long); if (mem / sizeof(unsigned long) != wear_eb_count) { NS_ERR("Too many erase blocks for wear reporting\n"); -- cgit v0.10.2 From 5c53d819c71c63fdc91f30a59164583f68e2d63a Mon Sep 17 00:00:00 2001 From: liu chuansheng Date: Fri, 6 Jul 2012 09:50:08 -0700 Subject: printk: replacing the raw_spin_lock/unlock with raw_spin_lock/unlock_irq In function devkmsg_read/writev/llseek/poll/open()..., the function raw_spin_lock/unlock is used, there is potential deadlock case happening. CPU1: thread1 doing the cat /dev/kmsg: raw_spin_lock(&logbuf_lock); while (user->seq == log_next_seq) { when thread1 run here, at this time one interrupt is coming on CPU1 and running based on this thread,if the interrupt handle called the printk which need the logbuf_lock spin also, it will cause deadlock. So we should use raw_spin_lock/unlock_irq here. Acked-by: Kay Sievers Signed-off-by: liu chuansheng Signed-off-by: Greg Kroah-Hartman diff --git a/kernel/printk.c b/kernel/printk.c index dba1821..12886cd 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -430,20 +430,20 @@ static ssize_t devkmsg_read(struct file *file, char __user *buf, ret = mutex_lock_interruptible(&user->lock); if (ret) return ret; - raw_spin_lock(&logbuf_lock); + raw_spin_lock_irq(&logbuf_lock); while (user->seq == log_next_seq) { if (file->f_flags & O_NONBLOCK) { ret = -EAGAIN; - raw_spin_unlock(&logbuf_lock); + raw_spin_unlock_irq(&logbuf_lock); goto out; } - raw_spin_unlock(&logbuf_lock); + raw_spin_unlock_irq(&logbuf_lock); ret = wait_event_interruptible(log_wait, user->seq != log_next_seq); if (ret) goto out; - raw_spin_lock(&logbuf_lock); + raw_spin_lock_irq(&logbuf_lock); } if (user->seq < log_first_seq) { @@ -451,7 +451,7 @@ static ssize_t devkmsg_read(struct file *file, char __user *buf, user->idx = log_first_idx; user->seq = log_first_seq; ret = -EPIPE; - raw_spin_unlock(&logbuf_lock); + raw_spin_unlock_irq(&logbuf_lock); goto out; } @@ -501,7 +501,7 @@ static ssize_t devkmsg_read(struct file *file, char __user *buf, user->idx = log_next(user->idx); user->seq++; - raw_spin_unlock(&logbuf_lock); + raw_spin_unlock_irq(&logbuf_lock); if (len > count) { ret = -EINVAL; @@ -528,7 +528,7 @@ static loff_t devkmsg_llseek(struct file *file, loff_t offset, int whence) if (offset) return -ESPIPE; - raw_spin_lock(&logbuf_lock); + raw_spin_lock_irq(&logbuf_lock); switch (whence) { case SEEK_SET: /* the first record */ @@ -552,7 +552,7 @@ static loff_t devkmsg_llseek(struct file *file, loff_t offset, int whence) default: ret = -EINVAL; } - raw_spin_unlock(&logbuf_lock); + raw_spin_unlock_irq(&logbuf_lock); return ret; } @@ -566,14 +566,14 @@ static unsigned int devkmsg_poll(struct file *file, poll_table *wait) poll_wait(file, &log_wait, wait); - raw_spin_lock(&logbuf_lock); + raw_spin_lock_irq(&logbuf_lock); if (user->seq < log_next_seq) { /* return error when data has vanished underneath us */ if (user->seq < log_first_seq) ret = POLLIN|POLLRDNORM|POLLERR|POLLPRI; ret = POLLIN|POLLRDNORM; } - raw_spin_unlock(&logbuf_lock); + raw_spin_unlock_irq(&logbuf_lock); return ret; } @@ -597,10 +597,10 @@ static int devkmsg_open(struct inode *inode, struct file *file) mutex_init(&user->lock); - raw_spin_lock(&logbuf_lock); + raw_spin_lock_irq(&logbuf_lock); user->idx = log_first_idx; user->seq = log_first_seq; - raw_spin_unlock(&logbuf_lock); + raw_spin_unlock_irq(&logbuf_lock); file->private_data = user; return 0; -- cgit v0.10.2 From e3f5a5f27153228569f3396049838e9727dae86e Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Fri, 6 Jul 2012 09:50:09 -0700 Subject: kmsg: escape the backslash character while exporting data Non-printable characters in the log data are hex-escaped to ensure safe post processing. We need to escape a backslash we find in the data, to be able to distinguish it from a backslash we add for the escaping. Also escape the non-printable character 127. Thanks to Miloslav Trmac for the heads up. Reported-by: Michael Neuling Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman diff --git a/kernel/printk.c b/kernel/printk.c index 12886cd..505863a 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -465,7 +465,7 @@ static ssize_t devkmsg_read(struct file *file, char __user *buf, for (i = 0; i < msg->text_len; i++) { unsigned char c = log_text(msg)[i]; - if (c < ' ' || c >= 128) + if (c < ' ' || c >= 127 || c == '\\') len += sprintf(user->buf + len, "\\x%02x", c); else user->buf[len++] = c; @@ -489,7 +489,7 @@ static ssize_t devkmsg_read(struct file *file, char __user *buf, continue; } - if (c < ' ' || c >= 128) { + if (c < ' ' || c >= 127 || c == '\\') { len += sprintf(user->buf + len, "\\x%02x", c); continue; } -- cgit v0.10.2 From 43a73a50b352cd3df25b3ced72033942a6a0f919 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Fri, 6 Jul 2012 09:50:09 -0700 Subject: kmsg: add the facility number to the syslog prefix After the recent split of facility and level into separate variables, we miss the facility value (always 0 for kernel-originated messages) in the syslog prefix. On Tue, Jul 3, 2012 at 12:45 PM, Dan Carpenter wrote: > Static checkers complain about the impossible condition here. > > In 084681d14e ('printk: flush continuation lines immediately to > console'), we changed msg->level from being a u16 to being an unsigned > 3 bit bitfield. Cc: Dan Carpenter Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman diff --git a/kernel/printk.c b/kernel/printk.c index 505863a..37cde75 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -818,15 +818,18 @@ static size_t print_time(u64 ts, char *buf) static size_t print_prefix(const struct log *msg, bool syslog, char *buf) { size_t len = 0; + unsigned int prefix = (msg->facility << 3) | msg->level; if (syslog) { if (buf) { - len += sprintf(buf, "<%u>", msg->level); + len += sprintf(buf, "<%u>", prefix); } else { len += 3; - if (msg->level > 9) - len++; - if (msg->level > 99) + if (prefix > 999) + len += 3; + else if (prefix > 99) + len += 2; + else if (prefix > 9) len++; } } -- cgit v0.10.2 From cb424ffe9f45ad80267f2a98fbd9bf21caa0ce22 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Fri, 6 Jul 2012 09:50:09 -0700 Subject: kmsg: properly handle concurrent non-blocking read() from /proc/kmsg The /proc/kmsg read() interface is internally simply wired up to a sequence of syslog() syscalls, which might are racy between their checks and actions, regarding concurrency. In the (very uncommon) case of concurrent readers of /dev/kmsg, relying on usual O_NONBLOCK behavior, the recently introduced mutex might block an O_NONBLOCK reader in read(), when poll() returns for it, but another process has already read the data in the meantime. We've seen that while running artificial test setups and tools that "fight" about /proc/kmsg data. This restores the original /proc/kmsg behavior, where in case of concurrent read()s, poll() might wake up but the read() syscall will just return 0 to the caller, while another process has "stolen" the data. This is in the general case not the expected behavior, but it is the exact same one, that can easily be triggered with a 3.4 kernel, and some tools might just rely on it. The mutex is not needed, the original integrity issue which introduced it, is in the meantime covered by: "fill buffer with more than a single message for SYSLOG_ACTION_READ" 116e90b23f74d303e8d607c7a7d54f60f14ab9f2 Cc: Yuanhan Liu Acked-by: Jan Beulich Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman diff --git a/kernel/printk.c b/kernel/printk.c index 37cde75..be9a82b 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -1021,7 +1021,6 @@ int do_syslog(int type, char __user *buf, int len, bool from_file) { bool clear = false; static int saved_console_loglevel = -1; - static DEFINE_MUTEX(syslog_mutex); int error; error = check_syslog_permissions(type, from_file); @@ -1048,17 +1047,11 @@ int do_syslog(int type, char __user *buf, int len, bool from_file) error = -EFAULT; goto out; } - error = mutex_lock_interruptible(&syslog_mutex); - if (error) - goto out; error = wait_event_interruptible(log_wait, syslog_seq != log_next_seq); - if (error) { - mutex_unlock(&syslog_mutex); + if (error) goto out; - } error = syslog_print(buf, len); - mutex_unlock(&syslog_mutex); break; /* Read/clear last kernel messages */ case SYSLOG_ACTION_READ_CLEAR: -- cgit v0.10.2 From 68b6507dc554ba015b5ed5e13b1ed4993cdf4024 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Fri, 6 Jul 2012 09:50:09 -0700 Subject: kmsg: make sure all messages reach a newly registered boot console We suppress printing kmsg records to the console, which are already printed immediately while we have received their fragments. Newly registered boot consoles print the entire kmsg buffer during registration. Clear the console-suppress flag after we skipped the record during its first storage, so any later print will see these records as usual. Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman diff --git a/kernel/printk.c b/kernel/printk.c index be9a82b..f02f1f5 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -1953,6 +1953,12 @@ skip: */ console_idx = log_next(console_idx); console_seq++; + /* + * We will get here again when we register a new + * CON_PRINTBUFFER console. Clear the flag so we + * will properly dump everything later. + */ + msg->flags &= ~LOG_NOCONS; goto skip; } -- cgit v0.10.2 From 9ab4233dd08036fe34a89c7dc6f47a8bf2eb29eb Mon Sep 17 00:00:00 2001 From: Andy Lutomirski Date: Thu, 5 Jul 2012 16:00:11 -0700 Subject: mm: Hold a file reference in madvise_remove Otherwise the code races with munmap (causing a use-after-free of the vma) or with close (causing a use-after-free of the struct file). The bug was introduced by commit 90ed52ebe481 ("[PATCH] holepunch: fix mmap_sem i_mutex deadlock") Cc: Hugh Dickins Cc: Miklos Szeredi Cc: Badari Pulavarty Cc: Nick Piggin Cc: stable@vger.kernel.org Signed-off-by: Andy Lutomirski Signed-off-by: Linus Torvalds diff --git a/mm/madvise.c b/mm/madvise.c index deff1b6..14d260f 100644 --- a/mm/madvise.c +++ b/mm/madvise.c @@ -15,6 +15,7 @@ #include #include #include +#include /* * Any behaviour which results in changes to the vma->vm_flags needs to @@ -204,14 +205,16 @@ static long madvise_remove(struct vm_area_struct *vma, { loff_t offset; int error; + struct file *f; *prev = NULL; /* tell sys_madvise we drop mmap_sem */ if (vma->vm_flags & (VM_LOCKED|VM_NONLINEAR|VM_HUGETLB)) return -EINVAL; - if (!vma->vm_file || !vma->vm_file->f_mapping - || !vma->vm_file->f_mapping->host) { + f = vma->vm_file; + + if (!f || !f->f_mapping || !f->f_mapping->host) { return -EINVAL; } @@ -221,11 +224,18 @@ static long madvise_remove(struct vm_area_struct *vma, offset = (loff_t)(start - vma->vm_start) + ((loff_t)vma->vm_pgoff << PAGE_SHIFT); - /* filesystem's fallocate may need to take i_mutex */ + /* + * Filesystem's fallocate may need to take i_mutex. We need to + * explicitly grab a reference because the vma (and hence the + * vma's reference to the file) can go away as soon as we drop + * mmap_sem. + */ + get_file(f); up_read(¤t->mm->mmap_sem); - error = do_fallocate(vma->vm_file, + error = do_fallocate(f, FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, offset, end - start); + fput(f); down_read(¤t->mm->mmap_sem); return error; } -- cgit v0.10.2 From 84774e615749acfce6b2f679b95924dffb5f5de1 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Thu, 5 Jul 2012 15:15:36 +0100 Subject: of: address: Don't fail a lookup just because a node has no reg property Sometimes it doesn't make any sense for a node to have an address. In this case device lookup will always be unsuccessful because we currently assume every node will have a reg property. This patch changes the semantics so that the resource address and the lookup address will only be compared if one exists. Things like AUXDATA() rely on of_dev_lookup to return the lookup entry of a particular device in order to do things like apply platform_data to a device. However, this is currently broken for nodes which do not have a reg property, meaning that platform_data can not be passed in those cases. Acked-by: Rob Herring Acked-by: Arnd Bergmann Signed-off-by: Lee Jones diff --git a/drivers/of/platform.c b/drivers/of/platform.c index 343ad29..9600480 100644 --- a/drivers/of/platform.c +++ b/drivers/of/platform.c @@ -317,10 +317,9 @@ static const struct of_dev_auxdata *of_dev_lookup(const struct of_dev_auxdata *l for(; lookup->compatible != NULL; lookup++) { if (!of_device_is_compatible(np, lookup->compatible)) continue; - if (of_address_to_resource(np, 0, &res)) - continue; - if (res.start != lookup->phys_addr) - continue; + if (!of_address_to_resource(np, 0, &res)) + if (res.start != lookup->phys_addr) + continue; pr_debug("%s: devname=%s\n", np->full_name, lookup->name); return lookup; } -- cgit v0.10.2 From c57920e6c23350b08c1b05606aafe356ebc6d8cc Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Fri, 15 Jun 2012 09:27:54 +0100 Subject: ARM: ux500: Fix build errors/warnings when MACH_UX500_DT is not set MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When MACH_UX500_DT and all related Device Tree configurations are forced off the warning and error below prevent the kernel from compiling. This simple patch fixes both issues and allows for full build and boot of ST-Ericsson's low-cost development board, Snowball. Warnings fixed: arch/arm/mach-ux500/board-mop500.c:680:32: warning: ‘snowball_of_platform_devs’ defined but not used Errors fixed: arch/arm/mach-ux500/timer.c: In function ‘ux500_timer_init’: arch/arm/mach-ux500/timer.c:66:3: error: implicit declaration of function ‘of_find_matching_node’ arch/arm/mach-ux500/timer.c:66:6: warning: assignment makes pointer from integer without a cast Cc: stable@vger.kernel.org Acked-by: Linus Walleij Signed-off-by: Lee Jones diff --git a/arch/arm/mach-ux500/board-mop500.c b/arch/arm/mach-ux500/board-mop500.c index 1509a3c..461012a 100644 --- a/arch/arm/mach-ux500/board-mop500.c +++ b/arch/arm/mach-ux500/board-mop500.c @@ -625,11 +625,6 @@ static struct platform_device *snowball_platform_devs[] __initdata = { &ab8500_device, }; -static struct platform_device *snowball_of_platform_devs[] __initdata = { - &snowball_led_dev, - &snowball_key_dev, -}; - static void __init mop500_init_machine(void) { struct device *parent = NULL; @@ -769,6 +764,11 @@ MACHINE_END #ifdef CONFIG_MACH_UX500_DT +static struct platform_device *snowball_of_platform_devs[] __initdata = { + &snowball_led_dev, + &snowball_key_dev, +}; + struct of_dev_auxdata u8500_auxdata_lookup[] __initdata = { /* Requires DMA and call-back bindings. */ OF_DEV_AUXDATA("arm,pl011", 0x80120000, "uart0", &uart0_plat), diff --git a/arch/arm/mach-ux500/timer.c b/arch/arm/mach-ux500/timer.c index 741e71f..66e7f00 100644 --- a/arch/arm/mach-ux500/timer.c +++ b/arch/arm/mach-ux500/timer.c @@ -63,8 +63,10 @@ static void __init ux500_timer_init(void) /* TODO: Once MTU has been DT:ed place code above into else. */ if (of_have_populated_dt()) { +#ifdef CONFIG_OF np = of_find_matching_node(NULL, prcmu_timer_of_match); if (!np) +#endif goto dt_fail; tmp_base = of_iomap(np, 0); -- cgit v0.10.2 From 2b667a2d8005e7c8362a77365cedbcd71fa5d6c1 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Fri, 6 Jul 2012 15:59:01 +0100 Subject: ARM: ux500: Over-ride the DT device naming scheme for pinctrl When pin control mapping tables are written the registered device name is supplied for use in name-based searches within the pinctrl driver. In the case of the DB8500 the string "pinctrl-db8500" is used. However, when we register the driver with Device Tree, its naming convention uses something that looks more like "pinctrl.2". To work around the device naming inconsistencies between devices registered via platform code and the ones registered by Device Tree, we use AUXDATA to over-ride the Device Tree naming scheme. Acked-by: Linus Walleij Signed-off-by: Lee Jones diff --git a/arch/arm/mach-ux500/board-mop500.c b/arch/arm/mach-ux500/board-mop500.c index 461012a..4fd93f5 100644 --- a/arch/arm/mach-ux500/board-mop500.c +++ b/arch/arm/mach-ux500/board-mop500.c @@ -786,6 +786,8 @@ struct of_dev_auxdata u8500_auxdata_lookup[] __initdata = { OF_DEV_AUXDATA("st,nomadik-gpio", 0x8011e000, "gpio.6", NULL), OF_DEV_AUXDATA("st,nomadik-gpio", 0x8011e080, "gpio.7", NULL), OF_DEV_AUXDATA("st,nomadik-gpio", 0xa03fe000, "gpio.8", NULL), + /* Requires device name bindings. */ + OF_DEV_AUXDATA("stericsson,nmk_pinctrl", 0, "pinctrl-db8500", NULL), {}, }; -- cgit v0.10.2 From 863555be0c81558b1af277addcf68acb8f778860 Mon Sep 17 00:00:00 2001 From: Mark Rustad Date: Tue, 26 Jun 2012 15:57:30 -0700 Subject: tcm_fc: Resolve suspicious RCU usage warnings Use rcu_dereference_protected to tell rcu that the ft_lport_lock is held during ft_lport_create. This resolved "suspicious RCU usage" warnings when debugging options are turned on. Signed-off-by: Mark Rustad Tested-by: Ross Brattain Cc: Signed-off-by: Nicholas Bellinger diff --git a/drivers/target/tcm_fc/tfc_sess.c b/drivers/target/tcm_fc/tfc_sess.c index cb99da9..87901fa 100644 --- a/drivers/target/tcm_fc/tfc_sess.c +++ b/drivers/target/tcm_fc/tfc_sess.c @@ -58,7 +58,8 @@ static struct ft_tport *ft_tport_create(struct fc_lport *lport) struct ft_tport *tport; int i; - tport = rcu_dereference(lport->prov[FC_TYPE_FCP]); + tport = rcu_dereference_protected(lport->prov[FC_TYPE_FCP], + lockdep_is_held(&ft_lport_lock)); if (tport && tport->tpg) return tport; -- cgit v0.10.2 From 873e9f7a3b8a5ca36085437da364654bfe3e5974 Mon Sep 17 00:00:00 2001 From: Magnus Damm Date: Fri, 6 Jul 2012 17:20:05 +0900 Subject: ARM: shmobile: fix platsmp.c build when ARCH_SH73A0=n Fix build error in the case of SMP=y but ARCH_SH73A0=n introduced by: 9601e87 ARM: shmobile: fix smp build The use of of_machine_is_compatible() will link in the the SoC-specific symbols: "sh73a0_get_core_count", "sh73a0_smp_prepare_cpus", "sh73a0_secondary_init" and "sh73a0_boot_secondary". This patch adds an ugly #ifdef wrapper as a stop-gap solution. Signed-off-by: Magnus Damm Tested-by: Tested-by: Simon Horman Acked-by: "Rafael J. Wysocki" Signed-off-by: Arnd Bergmann diff --git a/arch/arm/mach-shmobile/platsmp.c b/arch/arm/mach-shmobile/platsmp.c index e859fcd..fde0d23 100644 --- a/arch/arm/mach-shmobile/platsmp.c +++ b/arch/arm/mach-shmobile/platsmp.c @@ -22,8 +22,13 @@ #include #include +#ifdef CONFIG_ARCH_SH73A0 #define is_sh73a0() (machine_is_ag5evm() || machine_is_kota2() || \ of_machine_is_compatible("renesas,sh73a0")) +#else +#define is_sh73a0() (0) +#endif + #define is_r8a7779() machine_is_marzen() #ifdef CONFIG_ARCH_EMEV2 -- cgit v0.10.2 From 4f1d0f1971ba37010603a3a5c763f398b888d2f4 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Wed, 4 Jul 2012 16:35:35 +0100 Subject: qla2xxx: print the right array elements in qlt_async_event Based upon Alan's patch from Coverity scan id 793583, these debug messages in qlt_async_event() should be starting from byte 0, which is always the Asynchronous Event Status Code from the parent switch statement. Also, rename reason_code -> login_code following the language used in 2500 FW spec for Port Database Changed (0x8014) -> Port Database Changed Event Mailbox Register for mailbox[2]. Signed-off-by: Alan Cox Cc: Chad Dupuis Cc: Giridhar Malavali Signed-off-by: Nicholas Bellinger diff --git a/drivers/scsi/qla2xxx/qla_target.c b/drivers/scsi/qla2xxx/qla_target.c index 6986552..77759c7 100644 --- a/drivers/scsi/qla2xxx/qla_target.c +++ b/drivers/scsi/qla2xxx/qla_target.c @@ -3960,7 +3960,7 @@ void qlt_async_event(uint16_t code, struct scsi_qla_host *vha, { struct qla_hw_data *ha = vha->hw; struct qla_tgt *tgt = ha->tgt.qla_tgt; - int reason_code; + int login_code; ql_dbg(ql_dbg_tgt, vha, 0xe039, "scsi(%ld): ha state %d init_done %d oper_mode %d topo %d\n", @@ -4003,9 +4003,9 @@ void qlt_async_event(uint16_t code, struct scsi_qla_host *vha, { ql_dbg(ql_dbg_tgt_mgt, vha, 0xf03b, "qla_target(%d): Async LOOP_UP occured " - "(m[1]=%x, m[2]=%x, m[3]=%x, m[4]=%x)", vha->vp_idx, - le16_to_cpu(mailbox[1]), le16_to_cpu(mailbox[2]), - le16_to_cpu(mailbox[3]), le16_to_cpu(mailbox[4])); + "(m[0]=%x, m[1]=%x, m[2]=%x, m[3]=%x)", vha->vp_idx, + le16_to_cpu(mailbox[0]), le16_to_cpu(mailbox[1]), + le16_to_cpu(mailbox[2]), le16_to_cpu(mailbox[3])); if (tgt->link_reinit_iocb_pending) { qlt_send_notify_ack(vha, (void *)&tgt->link_reinit_iocb, 0, 0, 0, 0, 0, 0); @@ -4020,23 +4020,24 @@ void qlt_async_event(uint16_t code, struct scsi_qla_host *vha, case MBA_RSCN_UPDATE: ql_dbg(ql_dbg_tgt_mgt, vha, 0xf03c, "qla_target(%d): Async event %#x occured " - "(m[1]=%x, m[2]=%x, m[3]=%x, m[4]=%x)", vha->vp_idx, code, - le16_to_cpu(mailbox[1]), le16_to_cpu(mailbox[2]), - le16_to_cpu(mailbox[3]), le16_to_cpu(mailbox[4])); + "(m[0]=%x, m[1]=%x, m[2]=%x, m[3]=%x)", vha->vp_idx, code, + le16_to_cpu(mailbox[0]), le16_to_cpu(mailbox[1]), + le16_to_cpu(mailbox[2]), le16_to_cpu(mailbox[3])); break; case MBA_PORT_UPDATE: ql_dbg(ql_dbg_tgt_mgt, vha, 0xf03d, "qla_target(%d): Port update async event %#x " - "occured: updating the ports database (m[1]=%x, m[2]=%x, " - "m[3]=%x, m[4]=%x)", vha->vp_idx, code, - le16_to_cpu(mailbox[1]), le16_to_cpu(mailbox[2]), - le16_to_cpu(mailbox[3]), le16_to_cpu(mailbox[4])); - reason_code = le16_to_cpu(mailbox[2]); - if (reason_code == 0x4) + "occured: updating the ports database (m[0]=%x, m[1]=%x, " + "m[2]=%x, m[3]=%x)", vha->vp_idx, code, + le16_to_cpu(mailbox[0]), le16_to_cpu(mailbox[1]), + le16_to_cpu(mailbox[2]), le16_to_cpu(mailbox[3])); + + login_code = le16_to_cpu(mailbox[2]); + if (login_code == 0x4) ql_dbg(ql_dbg_tgt_mgt, vha, 0xf03e, "Async MB 2: Got PLOGI Complete\n"); - else if (reason_code == 0x7) + else if (login_code == 0x7) ql_dbg(ql_dbg_tgt_mgt, vha, 0xf03f, "Async MB 2: Port Logged Out\n"); break; @@ -4044,9 +4045,9 @@ void qlt_async_event(uint16_t code, struct scsi_qla_host *vha, default: ql_dbg(ql_dbg_tgt_mgt, vha, 0xf040, "qla_target(%d): Async event %#x occured: " - "ignore (m[1]=%x, m[2]=%x, m[3]=%x, m[4]=%x)", vha->vp_idx, - code, le16_to_cpu(mailbox[1]), le16_to_cpu(mailbox[2]), - le16_to_cpu(mailbox[3]), le16_to_cpu(mailbox[4])); + "ignore (m[0]=%x, m[1]=%x, m[2]=%x, m[3]=%x)", vha->vp_idx, + code, le16_to_cpu(mailbox[0]), le16_to_cpu(mailbox[1]), + le16_to_cpu(mailbox[2]), le16_to_cpu(mailbox[3])); break; } -- cgit v0.10.2 From 8dc6780587c99286c0d3de747a2946a76989414a Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Mon, 11 Jun 2012 09:24:11 -0700 Subject: eCryptfs: Gracefully refuse miscdev file ops on inherited/passed files File operations on /dev/ecryptfs would BUG() when the operations were performed by processes other than the process that originally opened the file. This could happen with open files inherited after fork() or file descriptors passed through IPC mechanisms. Rather than calling BUG(), an error code can be safely returned in most situations. In ecryptfs_miscdev_release(), eCryptfs still needs to handle the release even if the last file reference is being held by a process that didn't originally open the file. ecryptfs_find_daemon_by_euid() will not be successful, so a pointer to the daemon is stored in the file's private_data. The private_data pointer is initialized when the miscdev file is opened and only used when the file is released. https://launchpad.net/bugs/994247 Signed-off-by: Tyler Hicks Reported-by: Sasha Levin Tested-by: Sasha Levin diff --git a/fs/ecryptfs/miscdev.c b/fs/ecryptfs/miscdev.c index 3c632ec..c0038f6 100644 --- a/fs/ecryptfs/miscdev.c +++ b/fs/ecryptfs/miscdev.c @@ -49,7 +49,10 @@ ecryptfs_miscdev_poll(struct file *file, poll_table *pt) mutex_lock(&ecryptfs_daemon_hash_mux); /* TODO: Just use file->private_data? */ rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); - BUG_ON(rc || !daemon); + if (rc || !daemon) { + mutex_unlock(&ecryptfs_daemon_hash_mux); + return -EINVAL; + } mutex_lock(&daemon->mux); mutex_unlock(&ecryptfs_daemon_hash_mux); if (daemon->flags & ECRYPTFS_DAEMON_ZOMBIE) { @@ -122,6 +125,7 @@ ecryptfs_miscdev_open(struct inode *inode, struct file *file) goto out_unlock_daemon; } daemon->flags |= ECRYPTFS_DAEMON_MISCDEV_OPEN; + file->private_data = daemon; atomic_inc(&ecryptfs_num_miscdev_opens); out_unlock_daemon: mutex_unlock(&daemon->mux); @@ -152,9 +156,9 @@ ecryptfs_miscdev_release(struct inode *inode, struct file *file) mutex_lock(&ecryptfs_daemon_hash_mux); rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); - BUG_ON(rc || !daemon); + if (rc || !daemon) + daemon = file->private_data; mutex_lock(&daemon->mux); - BUG_ON(daemon->pid != task_pid(current)); BUG_ON(!(daemon->flags & ECRYPTFS_DAEMON_MISCDEV_OPEN)); daemon->flags &= ~ECRYPTFS_DAEMON_MISCDEV_OPEN; atomic_dec(&ecryptfs_num_miscdev_opens); @@ -270,8 +274,16 @@ ecryptfs_miscdev_read(struct file *file, char __user *buf, size_t count, mutex_lock(&ecryptfs_daemon_hash_mux); /* TODO: Just use file->private_data? */ rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); - BUG_ON(rc || !daemon); + if (rc || !daemon) { + mutex_unlock(&ecryptfs_daemon_hash_mux); + return -EINVAL; + } mutex_lock(&daemon->mux); + if (task_pid(current) != daemon->pid) { + mutex_unlock(&daemon->mux); + mutex_unlock(&ecryptfs_daemon_hash_mux); + return -EPERM; + } if (daemon->flags & ECRYPTFS_DAEMON_ZOMBIE) { rc = 0; mutex_unlock(&ecryptfs_daemon_hash_mux); @@ -308,9 +320,6 @@ check_list: * message from the queue; try again */ goto check_list; } - BUG_ON(euid != daemon->euid); - BUG_ON(current_user_ns() != daemon->user_ns); - BUG_ON(task_pid(current) != daemon->pid); msg_ctx = list_first_entry(&daemon->msg_ctx_out_queue, struct ecryptfs_msg_ctx, daemon_out_list); BUG_ON(!msg_ctx); -- cgit v0.10.2 From 82163edcdfa4eb3d74516cc8e9f38dd3d039b67d Mon Sep 17 00:00:00 2001 From: Santosh Nayak Date: Sat, 23 Jun 2012 07:59:54 -0300 Subject: [media] dvb-core: Release semaphore on error path dvb_register_device() There is a missing "up_write()" here. Semaphore should be released before returning error value. Cc: stable@kernel.org Signed-off-by: Santosh Nayak Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/dvb/dvb-core/dvbdev.c b/drivers/media/dvb/dvb-core/dvbdev.c index 00a6732..39eab73 100644 --- a/drivers/media/dvb/dvb-core/dvbdev.c +++ b/drivers/media/dvb/dvb-core/dvbdev.c @@ -243,6 +243,7 @@ int dvb_register_device(struct dvb_adapter *adap, struct dvb_device **pdvbdev, if (minor == MAX_DVB_MINORS) { kfree(dvbdevfops); kfree(dvbdev); + up_write(&minor_rwsem); mutex_unlock(&dvbdev_register_lock); return -EINVAL; } -- cgit v0.10.2 From 4dab0e5fe8ceaaf267003b5d7aa0c31c1092bee0 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Thu, 5 Jul 2012 22:52:34 -0300 Subject: [media] em28xx: fix em28xx-rc load The logic that checks if a device has remote control is wrong. Due to that, the em28xx RC module is not loaded by default. Fix the logic, in order to make it work properly. Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/em28xx/em28xx-cards.c b/drivers/media/video/em28xx/em28xx-cards.c index 92da7c2..862c657 100644 --- a/drivers/media/video/em28xx/em28xx-cards.c +++ b/drivers/media/video/em28xx/em28xx-cards.c @@ -2893,7 +2893,7 @@ static void request_module_async(struct work_struct *work) if (dev->board.has_dvb) request_module("em28xx-dvb"); - if (dev->board.has_ir_i2c && !disable_ir) + if (dev->board.ir_codes && !disable_ir) request_module("em28xx-rc"); } -- cgit v0.10.2 From a7deca6fa79d5c65575532e780f3c93f6bf8ddad Mon Sep 17 00:00:00 2001 From: David Dillow Date: Mon, 18 Jun 2012 00:15:21 -0300 Subject: [media] cx231xx: don't DMA to random addresses Commit 7a6f6c29d264cdd2fe0eb3d923217eed5f0ad134 (cx231xx: use URB_NO_TRANSFER_DMA_MAP) was intended to avoid mapping the DMA buffer for URB twice. This works for the URBs allocated with usb_alloc_urb(), as those are allocated from cohernent DMA pools, but the flag was also added for the VBI and audio URBs, which have a manually allocated area. This leaves the random trash in the structure after allocation as the DMA address, corrupting memory and preventing VBI and audio from working. Letting the USB core map the buffers solves the problem. Cc: stable@kernel.org Signed-off-by: David Dillow Cc: Sri Deevi Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx231xx/cx231xx-audio.c b/drivers/media/video/cx231xx/cx231xx-audio.c index 068f78d..b4c99c7 100644 --- a/drivers/media/video/cx231xx/cx231xx-audio.c +++ b/drivers/media/video/cx231xx/cx231xx-audio.c @@ -307,7 +307,7 @@ static int cx231xx_init_audio_isoc(struct cx231xx *dev) urb->context = dev; urb->pipe = usb_rcvisocpipe(dev->udev, dev->adev.end_point_addr); - urb->transfer_flags = URB_ISO_ASAP | URB_NO_TRANSFER_DMA_MAP; + urb->transfer_flags = URB_ISO_ASAP; urb->transfer_buffer = dev->adev.transfer_buffer[i]; urb->interval = 1; urb->complete = cx231xx_audio_isocirq; @@ -368,7 +368,7 @@ static int cx231xx_init_audio_bulk(struct cx231xx *dev) urb->context = dev; urb->pipe = usb_rcvbulkpipe(dev->udev, dev->adev.end_point_addr); - urb->transfer_flags = URB_NO_TRANSFER_DMA_MAP; + urb->transfer_flags = 0; urb->transfer_buffer = dev->adev.transfer_buffer[i]; urb->complete = cx231xx_audio_bulkirq; urb->transfer_buffer_length = sb_size; diff --git a/drivers/media/video/cx231xx/cx231xx-vbi.c b/drivers/media/video/cx231xx/cx231xx-vbi.c index 3d15314..ac7db52 100644 --- a/drivers/media/video/cx231xx/cx231xx-vbi.c +++ b/drivers/media/video/cx231xx/cx231xx-vbi.c @@ -448,7 +448,7 @@ int cx231xx_init_vbi_isoc(struct cx231xx *dev, int max_packets, return -ENOMEM; } dev->vbi_mode.bulk_ctl.urb[i] = urb; - urb->transfer_flags = URB_NO_TRANSFER_DMA_MAP; + urb->transfer_flags = 0; dev->vbi_mode.bulk_ctl.transfer_buffer[i] = kzalloc(sb_size, GFP_KERNEL); -- cgit v0.10.2 From a65c3262a766311d4704a9ea29354480c5e8590d Mon Sep 17 00:00:00 2001 From: Kamil Debski Date: Tue, 26 Jun 2012 04:44:40 -0300 Subject: [media] s5p-mfc: Fixed setup of custom controls in decoder and encoder Fixed bugs in functions that initialize custom controls. Signed-off-by: Kamil Debski Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-mfc/s5p_mfc_dec.c b/drivers/media/video/s5p-mfc/s5p_mfc_dec.c index 4dd32fc..feea867 100644 --- a/drivers/media/video/s5p-mfc/s5p_mfc_dec.c +++ b/drivers/media/video/s5p-mfc/s5p_mfc_dec.c @@ -996,6 +996,7 @@ int s5p_mfc_dec_ctrls_setup(struct s5p_mfc_ctx *ctx) for (i = 0; i < NUM_CTRLS; i++) { if (IS_MFC51_PRIV(controls[i].id)) { + memset(&cfg, 0, sizeof(struct v4l2_ctrl_config)); cfg.ops = &s5p_mfc_dec_ctrl_ops; cfg.id = controls[i].id; cfg.min = controls[i].minimum; diff --git a/drivers/media/video/s5p-mfc/s5p_mfc_enc.c b/drivers/media/video/s5p-mfc/s5p_mfc_enc.c index 03d8334..158b789 100644 --- a/drivers/media/video/s5p-mfc/s5p_mfc_enc.c +++ b/drivers/media/video/s5p-mfc/s5p_mfc_enc.c @@ -1773,6 +1773,7 @@ int s5p_mfc_enc_ctrls_setup(struct s5p_mfc_ctx *ctx) } for (i = 0; i < NUM_CTRLS; i++) { if (IS_MFC51_PRIV(controls[i].id)) { + memset(&cfg, 0, sizeof(struct v4l2_ctrl_config)); cfg.ops = &s5p_mfc_enc_ctrl_ops; cfg.id = controls[i].id; cfg.min = controls[i].minimum; -- cgit v0.10.2 From ba50e7e16b86e25048290b682f7a4e09e069d321 Mon Sep 17 00:00:00 2001 From: Devin Heitmueller Date: Sun, 1 Jul 2012 16:15:09 -0300 Subject: [media] cx25840: fix regression in HVR-1800 analog support The refactoring of the cx25840 driver to support the cx23888 caused breakage with the existing support for cx23885/cx23887 analog support. Rework the routines such that the new code is only used for the 888. Validated with the following boards: HVR-1800 retail (0070:7801) HVR-1800 OEM (0070:7809) HVR_1850 retail (0070:8541) Thanks to Steven Toth and Hauppauge for loaning me various boards to regression test with. Reported-by: Jonathan Thanks-to: Steven Toth Signed-off-by: Devin Heitmueler Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx25840/cx25840-core.c b/drivers/media/video/cx25840/cx25840-core.c index fc1ff69..a82b704 100644 --- a/drivers/media/video/cx25840/cx25840-core.c +++ b/drivers/media/video/cx25840/cx25840-core.c @@ -84,7 +84,7 @@ MODULE_PARM_DESC(debug, "Debugging messages [0=Off (default) 1=On]"); /* ----------------------------------------------------------------------- */ -static void cx23885_std_setup(struct i2c_client *client); +static void cx23888_std_setup(struct i2c_client *client); int cx25840_write(struct i2c_client *client, u16 addr, u8 value) { @@ -638,10 +638,13 @@ static void cx23885_initialize(struct i2c_client *client) finish_wait(&state->fw_wait, &wait); destroy_workqueue(q); - /* Call the cx23885 specific std setup func, we no longer rely on + /* Call the cx23888 specific std setup func, we no longer rely on * the generic cx24840 func. */ - cx23885_std_setup(client); + if (is_cx23888(state)) + cx23888_std_setup(client); + else + cx25840_std_setup(client); /* (re)set input */ set_input(client, state->vid_input, state->aud_input); @@ -1298,8 +1301,8 @@ static int set_v4lstd(struct i2c_client *client) } cx25840_and_or(client, 0x400, ~0xf, fmt); cx25840_and_or(client, 0x403, ~0x3, pal_m); - if (is_cx2388x(state)) - cx23885_std_setup(client); + if (is_cx23888(state)) + cx23888_std_setup(client); else cx25840_std_setup(client); if (!is_cx2583x(state)) @@ -1782,8 +1785,8 @@ static int cx25840_s_video_routing(struct v4l2_subdev *sd, struct cx25840_state *state = to_state(sd); struct i2c_client *client = v4l2_get_subdevdata(sd); - if (is_cx2388x(state)) - cx23885_std_setup(client); + if (is_cx23888(state)) + cx23888_std_setup(client); return set_input(client, input, state->aud_input); } @@ -1794,8 +1797,8 @@ static int cx25840_s_audio_routing(struct v4l2_subdev *sd, struct cx25840_state *state = to_state(sd); struct i2c_client *client = v4l2_get_subdevdata(sd); - if (is_cx2388x(state)) - cx23885_std_setup(client); + if (is_cx23888(state)) + cx23888_std_setup(client); return set_input(client, state->vid_input, input); } @@ -4939,7 +4942,7 @@ void cx23885_dif_setup(struct i2c_client *client, u32 ifHz) } } -static void cx23885_std_setup(struct i2c_client *client) +static void cx23888_std_setup(struct i2c_client *client) { struct cx25840_state *state = to_state(i2c_get_clientdata(client)); v4l2_std_id std = state->std; -- cgit v0.10.2 From e6d0db1d47c0ac4d61c842a95988cbe8712df447 Mon Sep 17 00:00:00 2001 From: Devin Heitmueller Date: Sun, 1 Jul 2012 16:15:10 -0300 Subject: [media] cx25840: fix regression in analog support hue/saturation controls Fix regression in HVR-1800 analog support hue/saturation controls. The changes made for the cx23888 caused regressions in the analog support for cx23885/cx23887 based boards (partly due to changes in the locations of the hue/saturation controls). As a result the wrong registers were being overwritten. Add code to use the correct registers if it's a cx23888 Validated with the following boards: HVR-1800 retail (0070:7801) HVR-1800 OEM (0070:7809) HVR-1850 retail (0070:8541) Thanks to Steven Toth and Hauppauge for loaning me various boards to regression test with. Reported-by: Jonathan Thanks-to: Steven Toth Signed-off-by: Devin Heitmueler Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx25840/cx25840-core.c b/drivers/media/video/cx25840/cx25840-core.c index a82b704..e12b3b0 100644 --- a/drivers/media/video/cx25840/cx25840-core.c +++ b/drivers/media/video/cx25840/cx25840-core.c @@ -1106,9 +1106,23 @@ static int set_input(struct i2c_client *client, enum cx25840_video_input vid_inp cx25840_write4(client, 0x410, 0xffff0dbf); cx25840_write4(client, 0x414, 0x00137d03); - cx25840_write4(client, 0x418, 0x01008080); + + /* on the 887, 0x418 is HSCALE_CTRL, on the 888 it is + CHROMA_CTRL */ + if (is_cx23888(state)) + cx25840_write4(client, 0x418, 0x01008080); + else + cx25840_write4(client, 0x418, 0x01000000); + cx25840_write4(client, 0x41c, 0x00000000); - cx25840_write4(client, 0x420, 0x001c3e0f); + + /* on the 887, 0x420 is CHROMA_CTRL, on the 888 it is + CRUSH_CTRL */ + if (is_cx23888(state)) + cx25840_write4(client, 0x420, 0x001c3e0f); + else + cx25840_write4(client, 0x420, 0x001c8282); + cx25840_write4(client, 0x42c, 0x42600000); cx25840_write4(client, 0x430, 0x0000039b); cx25840_write4(client, 0x438, 0x00000000); @@ -1315,6 +1329,7 @@ static int set_v4lstd(struct i2c_client *client) static int cx25840_s_ctrl(struct v4l2_ctrl *ctrl) { struct v4l2_subdev *sd = to_sd(ctrl); + struct cx25840_state *state = to_state(sd); struct i2c_client *client = v4l2_get_subdevdata(sd); switch (ctrl->id) { @@ -1327,12 +1342,20 @@ static int cx25840_s_ctrl(struct v4l2_ctrl *ctrl) break; case V4L2_CID_SATURATION: - cx25840_write(client, 0x420, ctrl->val << 1); - cx25840_write(client, 0x421, ctrl->val << 1); + if (is_cx23888(state)) { + cx25840_write(client, 0x418, ctrl->val << 1); + cx25840_write(client, 0x419, ctrl->val << 1); + } else { + cx25840_write(client, 0x420, ctrl->val << 1); + cx25840_write(client, 0x421, ctrl->val << 1); + } break; case V4L2_CID_HUE: - cx25840_write(client, 0x422, ctrl->val); + if (is_cx23888(state)) + cx25840_write(client, 0x41a, ctrl->val); + else + cx25840_write(client, 0x422, ctrl->val); break; default: -- cgit v0.10.2 From d90133ec58d4fad822ff322557b6885c5b77f127 Mon Sep 17 00:00:00 2001 From: Devin Heitmueller Date: Sun, 1 Jul 2012 16:15:11 -0300 Subject: [media] cx25840: fix regression in HVR-1800 analog audio The refactoring of the cx25840 driver to support the cx23888 caused breakage with the existing support for cx23885/cx23887 analog audio support. Tweak the code so that it only uses the code if it really is a cx23888 instead of applying it to all cx2388x based devices. Validated with the following boards: HVR-1800 retail (0070:7801) HVR-1800 OEM (0070:7809) HVR_1850 retail (0070:8541) Thanks to Steven Toth and Hauppauge for loaning me various boards to regression test with. Reported-by: Jonathan Thanks-to: Steven Toth Signed-off-by: Devin Heitmueler Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx25840/cx25840-core.c b/drivers/media/video/cx25840/cx25840-core.c index e12b3b0..7dc7bb1 100644 --- a/drivers/media/video/cx25840/cx25840-core.c +++ b/drivers/media/video/cx25840/cx25840-core.c @@ -1250,7 +1250,7 @@ static int set_input(struct i2c_client *client, enum cx25840_video_input vid_inp cx25840_write4(client, 0x8d0, 0x1f063870); } - if (is_cx2388x(state)) { + if (is_cx23888(state)) { /* HVR1850 */ /* AUD_IO_CTRL - I2S Input, Parallel1*/ /* - Channel 1 src - Parallel1 (Merlin out) */ -- cgit v0.10.2 From b5c5c17babfaf83f274ca8f0a7660284882d8d4d Mon Sep 17 00:00:00 2001 From: Devin Heitmueller Date: Sun, 1 Jul 2012 16:15:12 -0300 Subject: [media] cx25840: fix vsrc/hsrc usage on cx23888 designs The location of the vsrc/hsrc registers moved in the cx23888, causing the s_mbus call to fail prematurely indicating that "720x480 is not a valid size". The function bailed out before many pertinent registers were set related to the scaler (causing unexpected results in video rendering when doing raw video capture). Use the correct registers for the cx23888. Validated with the following boards: HVR-1800 retail (0070:7801) HVR-1800 OEM (0070:7809) HVR-1850 retail (0070:8541) Thanks to Steven Toth and Hauppauge for loaning me various boards to regression test with. Reported-by: Jonathan Thanks-to: Steven Toth Signed-off-by: Devin Heitmueler Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx25840/cx25840-core.c b/drivers/media/video/cx25840/cx25840-core.c index 7dc7bb1..d8eac3e 100644 --- a/drivers/media/video/cx25840/cx25840-core.c +++ b/drivers/media/video/cx25840/cx25840-core.c @@ -1380,11 +1380,21 @@ static int cx25840_s_mbus_fmt(struct v4l2_subdev *sd, struct v4l2_mbus_framefmt fmt->field = V4L2_FIELD_INTERLACED; fmt->colorspace = V4L2_COLORSPACE_SMPTE170M; - Vsrc = (cx25840_read(client, 0x476) & 0x3f) << 4; - Vsrc |= (cx25840_read(client, 0x475) & 0xf0) >> 4; + if (is_cx23888(state)) { + Vsrc = (cx25840_read(client, 0x42a) & 0x3f) << 4; + Vsrc |= (cx25840_read(client, 0x429) & 0xf0) >> 4; + } else { + Vsrc = (cx25840_read(client, 0x476) & 0x3f) << 4; + Vsrc |= (cx25840_read(client, 0x475) & 0xf0) >> 4; + } - Hsrc = (cx25840_read(client, 0x472) & 0x3f) << 4; - Hsrc |= (cx25840_read(client, 0x471) & 0xf0) >> 4; + if (is_cx23888(state)) { + Hsrc = (cx25840_read(client, 0x426) & 0x3f) << 4; + Hsrc |= (cx25840_read(client, 0x425) & 0xf0) >> 4; + } else { + Hsrc = (cx25840_read(client, 0x472) & 0x3f) << 4; + Hsrc |= (cx25840_read(client, 0x471) & 0xf0) >> 4; + } Vlines = fmt->height + (is_50Hz ? 4 : 7); -- cgit v0.10.2 From d214ddc86807fb1995fd3400347a202826332710 Mon Sep 17 00:00:00 2001 From: Devin Heitmueller Date: Sun, 1 Jul 2012 16:15:13 -0300 Subject: [media] cx23885: make analog support work for HVR_1250 (cx23885 variant) The analog support in the cx23885 driver was completely broken for the HVR-1250. Add the necessary code. Note that this only implements analog for the composite and s-video inputs. The tuner input continues to be non-functional due to a lack of analog support in the mt2131 driver. Validated with the following boards: HVR-1250 (0070:7911) Thanks to Steven Toth and Hauppauge for loaning me various boards to regression test with. Thanks-to: Steven Toth Signed-off-by: Devin Heitmueler Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx23885/cx23885-cards.c b/drivers/media/video/cx23885/cx23885-cards.c index 13739e0..f8664f2 100644 --- a/drivers/media/video/cx23885/cx23885-cards.c +++ b/drivers/media/video/cx23885/cx23885-cards.c @@ -127,22 +127,37 @@ struct cx23885_board cx23885_boards[] = { }, [CX23885_BOARD_HAUPPAUGE_HVR1250] = { .name = "Hauppauge WinTV-HVR1250", + .porta = CX23885_ANALOG_VIDEO, .portc = CX23885_MPEG_DVB, +#ifdef MT2131_NO_ANALOG_SUPPORT_YET + .tuner_type = TUNER_PHILIPS_TDA8290, + .tuner_addr = 0x42, /* 0x84 >> 1 */ + .tuner_bus = 1, +#endif + .force_bff = 1, .input = {{ +#ifdef MT2131_NO_ANALOG_SUPPORT_YET .type = CX23885_VMUX_TELEVISION, - .vmux = 0, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN5_CH2 | + CX25840_VIN2_CH1, + .amux = CX25840_AUDIO8, .gpio0 = 0xff00, }, { - .type = CX23885_VMUX_DEBUG, - .vmux = 0, - .gpio0 = 0xff01, - }, { +#endif .type = CX23885_VMUX_COMPOSITE1, - .vmux = 1, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN4_CH2 | + CX25840_VIN6_CH1, + .amux = CX25840_AUDIO7, .gpio0 = 0xff02, }, { .type = CX23885_VMUX_SVIDEO, - .vmux = 2, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN4_CH2 | + CX25840_VIN8_CH1 | + CX25840_SVIDEO_ON, + .amux = CX25840_AUDIO7, .gpio0 = 0xff02, } }, }, @@ -1526,10 +1541,10 @@ void cx23885_card_setup(struct cx23885_dev *dev) */ switch (dev->board) { case CX23885_BOARD_TEVII_S470: - case CX23885_BOARD_HAUPPAUGE_HVR1250: /* Currently only enabled for the integrated IR controller */ if (!enable_885_ir) break; + case CX23885_BOARD_HAUPPAUGE_HVR1250: case CX23885_BOARD_HAUPPAUGE_HVR1800: case CX23885_BOARD_HAUPPAUGE_HVR1800lp: case CX23885_BOARD_HAUPPAUGE_HVR1700: diff --git a/drivers/media/video/cx23885/cx23885-video.c b/drivers/media/video/cx23885/cx23885-video.c index c654bdc..4d05689 100644 --- a/drivers/media/video/cx23885/cx23885-video.c +++ b/drivers/media/video/cx23885/cx23885-video.c @@ -505,6 +505,7 @@ static int cx23885_video_mux(struct cx23885_dev *dev, unsigned int input) if ((dev->board == CX23885_BOARD_HAUPPAUGE_HVR1800) || (dev->board == CX23885_BOARD_MPX885) || + (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1250) || (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1850)) { /* Configure audio routing */ v4l2_subdev_call(dev->sd_cx25840, audio, s_routing, -- cgit v0.10.2 From 0ac60acb5491df565141c0e3a87d7148a865fe36 Mon Sep 17 00:00:00 2001 From: Devin Heitmueller Date: Sun, 1 Jul 2012 16:15:14 -0300 Subject: [media] cx23885: add support for HVR-1255 analog (cx23888 variant) Get the HVR-1255 analog support working for all supported inputs. This includes introduction of a new board profile for an OEM variant which doesn't have all the same inputs as the retail version of the board. Validated with the following boards: HVR-1255 (0070:2259) Thanks to Steven Toth and Hauppauge for loaning me various boards to regression test with. Thanks-to: Steven Toth Signed-off-by: Devin Heitmueler Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx23885/cx23885-cards.c b/drivers/media/video/cx23885/cx23885-cards.c index f8664f2..bf79003 100644 --- a/drivers/media/video/cx23885/cx23885-cards.c +++ b/drivers/media/video/cx23885/cx23885-cards.c @@ -282,7 +282,55 @@ struct cx23885_board cx23885_boards[] = { }, [CX23885_BOARD_HAUPPAUGE_HVR1255] = { .name = "Hauppauge WinTV-HVR1255", + .porta = CX23885_ANALOG_VIDEO, .portc = CX23885_MPEG_DVB, + .tuner_type = TUNER_ABSENT, + .tuner_addr = 0x42, /* 0x84 >> 1 */ + .force_bff = 1, + .input = {{ + .type = CX23885_VMUX_TELEVISION, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN5_CH2 | + CX25840_VIN2_CH1 | + CX25840_DIF_ON, + .amux = CX25840_AUDIO8, + }, { + .type = CX23885_VMUX_COMPOSITE1, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN4_CH2 | + CX25840_VIN6_CH1, + .amux = CX25840_AUDIO7, + }, { + .type = CX23885_VMUX_SVIDEO, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN4_CH2 | + CX25840_VIN8_CH1 | + CX25840_SVIDEO_ON, + .amux = CX25840_AUDIO7, + } }, + }, + [CX23885_BOARD_HAUPPAUGE_HVR1255_22111] = { + .name = "Hauppauge WinTV-HVR1255", + .porta = CX23885_ANALOG_VIDEO, + .portc = CX23885_MPEG_DVB, + .tuner_type = TUNER_ABSENT, + .tuner_addr = 0x42, /* 0x84 >> 1 */ + .force_bff = 1, + .input = {{ + .type = CX23885_VMUX_TELEVISION, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN5_CH2 | + CX25840_VIN2_CH1 | + CX25840_DIF_ON, + .amux = CX25840_AUDIO8, + }, { + .type = CX23885_VMUX_SVIDEO, + .vmux = CX25840_VIN7_CH3 | + CX25840_VIN4_CH2 | + CX25840_VIN8_CH1 | + CX25840_SVIDEO_ON, + .amux = CX25840_AUDIO7, + } }, }, [CX23885_BOARD_HAUPPAUGE_HVR1210] = { .name = "Hauppauge WinTV-HVR1210", @@ -639,7 +687,7 @@ struct cx23885_subid cx23885_subids[] = { }, { .subvendor = 0x0070, .subdevice = 0x2259, - .card = CX23885_BOARD_HAUPPAUGE_HVR1255, + .card = CX23885_BOARD_HAUPPAUGE_HVR1255_22111, }, { .subvendor = 0x0070, .subdevice = 0x2291, @@ -1145,6 +1193,7 @@ void cx23885_gpio_setup(struct cx23885_dev *dev) case CX23885_BOARD_HAUPPAUGE_HVR1270: case CX23885_BOARD_HAUPPAUGE_HVR1275: case CX23885_BOARD_HAUPPAUGE_HVR1255: + case CX23885_BOARD_HAUPPAUGE_HVR1255_22111: case CX23885_BOARD_HAUPPAUGE_HVR1210: /* GPIO-5 RF Control: 0 = RF1 Terrestrial, 1 = RF2 Cable */ /* GPIO-6 I2C Gate which can isolate the demod from the bus */ @@ -1282,6 +1331,7 @@ int cx23885_ir_init(struct cx23885_dev *dev) case CX23885_BOARD_HAUPPAUGE_HVR1400: case CX23885_BOARD_HAUPPAUGE_HVR1275: case CX23885_BOARD_HAUPPAUGE_HVR1255: + case CX23885_BOARD_HAUPPAUGE_HVR1255_22111: case CX23885_BOARD_HAUPPAUGE_HVR1210: /* FIXME: Implement me */ break; @@ -1439,6 +1489,7 @@ void cx23885_card_setup(struct cx23885_dev *dev) case CX23885_BOARD_HAUPPAUGE_HVR1270: case CX23885_BOARD_HAUPPAUGE_HVR1275: case CX23885_BOARD_HAUPPAUGE_HVR1255: + case CX23885_BOARD_HAUPPAUGE_HVR1255_22111: case CX23885_BOARD_HAUPPAUGE_HVR1210: case CX23885_BOARD_HAUPPAUGE_HVR1850: case CX23885_BOARD_HAUPPAUGE_HVR1290: @@ -1526,6 +1577,7 @@ void cx23885_card_setup(struct cx23885_dev *dev) case CX23885_BOARD_HAUPPAUGE_HVR1270: case CX23885_BOARD_HAUPPAUGE_HVR1275: case CX23885_BOARD_HAUPPAUGE_HVR1255: + case CX23885_BOARD_HAUPPAUGE_HVR1255_22111: case CX23885_BOARD_HAUPPAUGE_HVR1210: case CX23885_BOARD_COMPRO_VIDEOMATE_E800: case CX23885_BOARD_HAUPPAUGE_HVR1290: @@ -1554,6 +1606,8 @@ void cx23885_card_setup(struct cx23885_dev *dev) case CX23885_BOARD_NETUP_DUAL_DVBS2_CI: case CX23885_BOARD_NETUP_DUAL_DVB_T_C_CI_RF: case CX23885_BOARD_COMPRO_VIDEOMATE_E800: + case CX23885_BOARD_HAUPPAUGE_HVR1255: + case CX23885_BOARD_HAUPPAUGE_HVR1255_22111: case CX23885_BOARD_HAUPPAUGE_HVR1270: case CX23885_BOARD_HAUPPAUGE_HVR1850: case CX23885_BOARD_MYGICA_X8506: diff --git a/drivers/media/video/cx23885/cx23885-dvb.c b/drivers/media/video/cx23885/cx23885-dvb.c index a80a92c..cd54268 100644 --- a/drivers/media/video/cx23885/cx23885-dvb.c +++ b/drivers/media/video/cx23885/cx23885-dvb.c @@ -712,6 +712,7 @@ static int dvb_register(struct cx23885_tsport *port) } break; case CX23885_BOARD_HAUPPAUGE_HVR1255: + case CX23885_BOARD_HAUPPAUGE_HVR1255_22111: i2c_bus = &dev->i2c_bus[0]; fe0->dvb.frontend = dvb_attach(s5h1411_attach, &hcw_s5h1411_config, @@ -721,6 +722,11 @@ static int dvb_register(struct cx23885_tsport *port) 0x60, &dev->i2c_bus[1].i2c_adap, &hauppauge_tda18271_config); } + + tda18271_attach(&dev->ts1.analog_fe, + 0x60, &dev->i2c_bus[1].i2c_adap, + &hauppauge_tda18271_config); + break; case CX23885_BOARD_HAUPPAUGE_HVR1800: i2c_bus = &dev->i2c_bus[0]; diff --git a/drivers/media/video/cx23885/cx23885-video.c b/drivers/media/video/cx23885/cx23885-video.c index 4d05689..22f8e7f 100644 --- a/drivers/media/video/cx23885/cx23885-video.c +++ b/drivers/media/video/cx23885/cx23885-video.c @@ -506,6 +506,8 @@ static int cx23885_video_mux(struct cx23885_dev *dev, unsigned int input) if ((dev->board == CX23885_BOARD_HAUPPAUGE_HVR1800) || (dev->board == CX23885_BOARD_MPX885) || (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1250) || + (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1255) || + (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1255_22111) || (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1850)) { /* Configure audio routing */ v4l2_subdev_call(dev->sd_cx25840, audio, s_routing, @@ -1579,7 +1581,9 @@ static int cx23885_set_freq_via_ops(struct cx23885_dev *dev, fe = vfe->dvb.frontend; - if (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1850) + if ((dev->board == CX23885_BOARD_HAUPPAUGE_HVR1850) || + (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1255) || + (dev->board == CX23885_BOARD_HAUPPAUGE_HVR1255_22111)) fe = &dev->ts1.analog_fe; if (fe && fe->ops.tuner_ops.set_analog_params) { @@ -1609,6 +1613,8 @@ int cx23885_set_frequency(struct file *file, void *priv, int ret; switch (dev->board) { + case CX23885_BOARD_HAUPPAUGE_HVR1255: + case CX23885_BOARD_HAUPPAUGE_HVR1255_22111: case CX23885_BOARD_HAUPPAUGE_HVR1850: ret = cx23885_set_freq_via_ops(dev, f); break; diff --git a/drivers/media/video/cx23885/cx23885.h b/drivers/media/video/cx23885/cx23885.h index d884784..13c37ec 100644 --- a/drivers/media/video/cx23885/cx23885.h +++ b/drivers/media/video/cx23885/cx23885.h @@ -90,6 +90,7 @@ #define CX23885_BOARD_MYGICA_X8507 33 #define CX23885_BOARD_TERRATEC_CINERGY_T_PCIE_DUAL 34 #define CX23885_BOARD_TEVII_S471 35 +#define CX23885_BOARD_HAUPPAUGE_HVR1255_22111 36 #define GPIO_0 0x00000001 #define GPIO_1 0x00000002 -- cgit v0.10.2 From c6cff169268bba9de687acf317ac24aa038cc263 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sun, 1 Jul 2012 21:38:03 -0300 Subject: [media] cx23885: Silence unknown command warnings I am seeing a constant stream of warnings on my cx23885 based card: cx23885_tuner_callback(): Unknown command 0x2. Add a check in cx23885_tuner_callback to silence it. Signed-off-by: Anton Blanchard Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/cx23885/cx23885-cards.c b/drivers/media/video/cx23885/cx23885-cards.c index bf79003..080e111 100644 --- a/drivers/media/video/cx23885/cx23885-cards.c +++ b/drivers/media/video/cx23885/cx23885-cards.c @@ -963,7 +963,7 @@ int cx23885_tuner_callback(void *priv, int component, int command, int arg) struct cx23885_dev *dev = port->dev; u32 bitmask = 0; - if (command == XC2028_RESET_CLK) + if ((command == XC2028_RESET_CLK) || (command == XC2028_I2C_FLUSH)) return 0; if (command != 0) { -- cgit v0.10.2 From 57f4422f7bd83bb0a092e3f9aea0d9c8ac59045e Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sun, 1 Jul 2012 21:58:00 -0300 Subject: [media] winbond-cir: Fix txandrx module info MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We aren't getting any module info for the txandx option because of a typo: parm: txandrx:bool Signed-off-by: Anton Blanchard Acked-by: David Härdeman Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/rc/winbond-cir.c b/drivers/media/rc/winbond-cir.c index 342c2c8..eca17b5 100644 --- a/drivers/media/rc/winbond-cir.c +++ b/drivers/media/rc/winbond-cir.c @@ -232,7 +232,7 @@ MODULE_PARM_DESC(invert, "Invert the signal from the IR receiver"); static bool txandrx; /* default = 0 */ module_param(txandrx, bool, 0444); -MODULE_PARM_DESC(invert, "Allow simultaneous TX and RX"); +MODULE_PARM_DESC(txandrx, "Allow simultaneous TX and RX"); static unsigned int wake_sc = 0x800F040C; module_param(wake_sc, uint, 0644); -- cgit v0.10.2 From 8299d62843d4dcc5d769ce901799b8e7806ec93f Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sun, 1 Jul 2012 21:58:52 -0300 Subject: [media] winbond-cir: Initialise timeout, driver_type and allowed_protos MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We need to set a timeout so we can go idle on no activity. Signed-off-by: Anton Blanchard Acked-by: David Härdeman Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/rc/winbond-cir.c b/drivers/media/rc/winbond-cir.c index eca17b5..54ee348 100644 --- a/drivers/media/rc/winbond-cir.c +++ b/drivers/media/rc/winbond-cir.c @@ -1032,6 +1032,8 @@ wbcir_probe(struct pnp_dev *device, const struct pnp_device_id *dev_id) data->dev->tx_ir = wbcir_tx; data->dev->priv = data; data->dev->dev.parent = &device->dev; + data->dev->timeout = MS_TO_NS(100); + data->dev->allowed_protos = RC_TYPE_ALL; if (!request_region(data->wbase, WAKEUP_IOMEM_LEN, DRVNAME)) { dev_err(dev, "Region 0x%lx-0x%lx already in use!\n", -- cgit v0.10.2 From 7103180b4385de908d69815dbd6807879d371cbf Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Thu, 5 Jul 2012 14:31:33 -0300 Subject: [media] omap3isp: preview: Fix output size computation depending on input format The preview engine crops 4 columns and 4 lines when CFA is enabled. Commit b2da46e52fe7871cba36e1a435844502c0eccf39 ("omap3isp: preview: Add support for greyscale input") inverted the condition by mistake, fix this. Reported-by: Florian Neuhaus Signed-off-by: Laurent Pinchart Tested-by: Florian Neuhaus Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/omap3isp/isppreview.c b/drivers/media/video/omap3isp/isppreview.c index 8a4935e..a48a747 100644 --- a/drivers/media/video/omap3isp/isppreview.c +++ b/drivers/media/video/omap3isp/isppreview.c @@ -1102,7 +1102,7 @@ static void preview_config_input_size(struct isp_prev_device *prev, u32 active) unsigned int elv = prev->crop.top + prev->crop.height - 1; u32 features; - if (format->code == V4L2_MBUS_FMT_Y10_1X10) { + if (format->code != V4L2_MBUS_FMT_Y10_1X10) { sph -= 2; eph += 2; slv -= 2; -- cgit v0.10.2 From 5aedc1094041335598c6aa73c5cf882f30886cd7 Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Fri, 6 Jul 2012 08:41:25 -0300 Subject: [media] omap3isp: preview: Fix contrast and brightness handling Commit bac387efbb88cf0e8df6f46a38387897cea464ee ("omap3isp: preview: Simplify configuration parameters access") added three fields to the preview_update structure, but failed to properly update the related initializers. Fix this. Signed-off-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/omap3isp/isppreview.c b/drivers/media/video/omap3isp/isppreview.c index a48a747..dd91da2 100644 --- a/drivers/media/video/omap3isp/isppreview.c +++ b/drivers/media/video/omap3isp/isppreview.c @@ -888,12 +888,12 @@ static const struct preview_update update_attrs[] = { preview_config_contrast, NULL, offsetof(struct prev_params, contrast), - 0, true, + 0, 0, true, }, /* OMAP3ISP_PREV_BRIGHTNESS */ { preview_config_brightness, NULL, offsetof(struct prev_params, brightness), - 0, true, + 0, 0, true, }, }; -- cgit v0.10.2 From 4e39da01027eb73556c1ad416945e37f2771464e Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Mon, 4 Jun 2012 13:15:56 -0300 Subject: [media] s5p-fimc: Add missing FIMC-LITE file operations locking commit 5126f2590bee412e3053de851cb07f531e4be36a "v4l2-dev: add flag to have the core lock all file operations" introduced an additional bit flag (V4L2_FL_LOCK_ALL_FOPS) that should be set by drivers that use the v4l2 core lock for all file operations. Since this driver has been merged at the same time as the core changes it doesn't set this flags and thus its all file operations except IOCTL are not properly serialized. Fix this by adding file ops locking in the driver. Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/s5p-fimc/fimc-lite.c b/drivers/media/video/s5p-fimc/fimc-lite.c index 4d269b8..74ff310 100644 --- a/drivers/media/video/s5p-fimc/fimc-lite.c +++ b/drivers/media/video/s5p-fimc/fimc-lite.c @@ -453,34 +453,42 @@ static int fimc_lite_open(struct file *file) struct fimc_lite *fimc = video_drvdata(file); int ret; + if (mutex_lock_interruptible(&fimc->lock)) + return -ERESTARTSYS; + set_bit(ST_FLITE_IN_USE, &fimc->state); ret = pm_runtime_get_sync(&fimc->pdev->dev); if (ret < 0) - return ret; - - if (++fimc->ref_count != 1 || fimc->out_path != FIMC_IO_DMA) - return 0; + goto done; ret = v4l2_fh_open(file); if (ret < 0) - return ret; + goto done; - ret = fimc_pipeline_initialize(&fimc->pipeline, &fimc->vfd->entity, - true); - if (ret < 0) { - pm_runtime_put_sync(&fimc->pdev->dev); - fimc->ref_count--; - v4l2_fh_release(file); - clear_bit(ST_FLITE_IN_USE, &fimc->state); - } + if (++fimc->ref_count == 1 && fimc->out_path == FIMC_IO_DMA) { + ret = fimc_pipeline_initialize(&fimc->pipeline, + &fimc->vfd->entity, true); + if (ret < 0) { + pm_runtime_put_sync(&fimc->pdev->dev); + fimc->ref_count--; + v4l2_fh_release(file); + clear_bit(ST_FLITE_IN_USE, &fimc->state); + } - fimc_lite_clear_event_counters(fimc); + fimc_lite_clear_event_counters(fimc); + } +done: + mutex_unlock(&fimc->lock); return ret; } static int fimc_lite_close(struct file *file) { struct fimc_lite *fimc = video_drvdata(file); + int ret; + + if (mutex_lock_interruptible(&fimc->lock)) + return -ERESTARTSYS; if (--fimc->ref_count == 0 && fimc->out_path == FIMC_IO_DMA) { clear_bit(ST_FLITE_IN_USE, &fimc->state); @@ -494,20 +502,39 @@ static int fimc_lite_close(struct file *file) if (fimc->ref_count == 0) vb2_queue_release(&fimc->vb_queue); - return v4l2_fh_release(file); + ret = v4l2_fh_release(file); + + mutex_unlock(&fimc->lock); + return ret; } static unsigned int fimc_lite_poll(struct file *file, struct poll_table_struct *wait) { struct fimc_lite *fimc = video_drvdata(file); - return vb2_poll(&fimc->vb_queue, file, wait); + int ret; + + if (mutex_lock_interruptible(&fimc->lock)) + return POLL_ERR; + + ret = vb2_poll(&fimc->vb_queue, file, wait); + mutex_unlock(&fimc->lock); + + return ret; } static int fimc_lite_mmap(struct file *file, struct vm_area_struct *vma) { struct fimc_lite *fimc = video_drvdata(file); - return vb2_mmap(&fimc->vb_queue, vma); + int ret; + + if (mutex_lock_interruptible(&fimc->lock)) + return -ERESTARTSYS; + + ret = vb2_mmap(&fimc->vb_queue, vma); + mutex_unlock(&fimc->lock); + + return ret; } static const struct v4l2_file_operations fimc_lite_fops = { -- cgit v0.10.2 From ec3ed85f926f4e900bc48cec6e72abbe6475321f Mon Sep 17 00:00:00 2001 From: Sylwester Nawrocki Date: Wed, 27 Jun 2012 10:12:31 -0300 Subject: [media] Revert "[media] V4L: JPEG class documentation corrections" This reverts commit feed0258e11e04b7e0, as the same issues are already covered in another version of that patch that was also applied (579e92ffac65c717c9c8a50feb755a). Signed-off-by: Sylwester Nawrocki Signed-off-by: Kyungmin Park Signed-off-by: Mauro Carvalho Chehab diff --git a/Documentation/DocBook/media/v4l/controls.xml b/Documentation/DocBook/media/v4l/controls.xml index 8994132..43ca749 100644 --- a/Documentation/DocBook/media/v4l/controls.xml +++ b/Documentation/DocBook/media/v4l/controls.xml @@ -3988,7 +3988,7 @@ interface and may change in the future. from RGB to Y'CbCr color space. - + diff --git a/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml b/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml index e3d5afcd..0a4b90f 100644 --- a/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml +++ b/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml @@ -284,13 +284,6 @@ These controls are described in . - - V4L2_CTRL_CLASS_JPEG - 0x9d0000 - The class containing JPEG compression controls. -These controls are described in . - -- cgit v0.10.2 From c540521bba5d2f24bd2c0417157bfaf8b85e2eee Mon Sep 17 00:00:00 2001 From: Andy Lutomirski Date: Thu, 5 Jul 2012 11:23:24 -0700 Subject: security: Minor improvements to no_new_privs documentation The documentation didn't actually mention how to enable no_new_privs. This also adds a note about possible interactions between no_new_privs and LSMs (i.e. why teaching systemd to set no_new_privs is not necessarily a good idea), and it references the new docs from include/linux/prctl.h. Suggested-by: Rob Landley Signed-off-by: Andy Lutomirski Acked-by: Kees Cook Signed-off-by: James Morris diff --git a/Documentation/prctl/no_new_privs.txt b/Documentation/prctl/no_new_privs.txt index cb705ec..f7be84f 100644 --- a/Documentation/prctl/no_new_privs.txt +++ b/Documentation/prctl/no_new_privs.txt @@ -25,6 +25,13 @@ bits will no longer change the uid or gid; file capabilities will not add to the permitted set, and LSMs will not relax constraints after execve. +To set no_new_privs, use prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0). + +Be careful, though: LSMs might also not tighten constraints on exec +in no_new_privs mode. (This means that setting up a general-purpose +service launcher to set no_new_privs before execing daemons may +interfere with LSM-based sandboxing.) + Note that no_new_privs does not prevent privilege changes that do not involve execve. An appropriately privileged task can still call setuid(2) and receive SCM_RIGHTS datagrams. diff --git a/include/linux/prctl.h b/include/linux/prctl.h index 3988012..289760f 100644 --- a/include/linux/prctl.h +++ b/include/linux/prctl.h @@ -141,6 +141,8 @@ * Changing LSM security domain is considered a new privilege. So, for example, * asking selinux for a specific new context (e.g. with runcon) will result * in execve returning -EPERM. + * + * See Documentation/prctl/no_new_privs.txt for more details. */ #define PR_SET_NO_NEW_PRIVS 38 #define PR_GET_NO_NEW_PRIVS 39 -- cgit v0.10.2 From 476a7eeb60e70ddab138e7cb4bc44ef5ac20782e Mon Sep 17 00:00:00 2001 From: Shinya Kuribayashi Date: Sat, 7 Jul 2012 13:37:42 +0300 Subject: hwspinlock/core: use global ID to register hwspinlocks on multiple devices Commit 300bab9770 (hwspinlock/core: register a bank of hwspinlocks in a single API call, 2011-09-06) introduced 'hwspin_lock_register_single()' to register numerous (a bank of) hwspinlock instances in a single API, 'hwspin_lock_register()'. At which time, 'hwspin_lock_register()' accidentally passes 'local IDs' to 'hwspin_lock_register_single()', despite that ..._single() requires 'global IDs' to register hwspinlocks. We have to convert into global IDs by supplying the missing 'base_id'. Cc: stable Signed-off-by: Shinya Kuribayashi [ohad: fix error path of hwspin_lock_register, too] Signed-off-by: Ohad Ben-Cohen diff --git a/drivers/hwspinlock/hwspinlock_core.c b/drivers/hwspinlock/hwspinlock_core.c index 61c9cf1..1201a15 100644 --- a/drivers/hwspinlock/hwspinlock_core.c +++ b/drivers/hwspinlock/hwspinlock_core.c @@ -345,7 +345,7 @@ int hwspin_lock_register(struct hwspinlock_device *bank, struct device *dev, spin_lock_init(&hwlock->lock); hwlock->bank = bank; - ret = hwspin_lock_register_single(hwlock, i); + ret = hwspin_lock_register_single(hwlock, base_id + i); if (ret) goto reg_failed; } @@ -354,7 +354,7 @@ int hwspin_lock_register(struct hwspinlock_device *bank, struct device *dev, reg_failed: while (--i >= 0) - hwspin_lock_unregister_single(i); + hwspin_lock_unregister_single(base_id + i); return ret; } EXPORT_SYMBOL_GPL(hwspin_lock_register); -- cgit v0.10.2 From 7db5b3ca0ecdb2e8fad52a4770e4e320e61c77a6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 7 Jul 2012 15:55:47 -0700 Subject: Revert "cgroup: superblock can't be released with active dentries" This reverts commit fa980ca87d15bb8a1317853f257a505990f3ffde. The commit was an attempt to fix a race condition where a cgroup hierarchy may be unmounted with positive dentry reference on root cgroup. While the commit made the race condition slightly more difficult to trigger, the race was still there and could be reliably triggered using a different test case. Revert the incorrect fix. The next commit will describe the race and fix it correctly. Signed-off-by: Tejun Heo LKML-Reference: <4FEEA5CB.8070809@huawei.com> Reported-by: shyju pv Cc: Sasha Levin Acked-by: Li Zefan diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 2097684..5f134a0 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -901,13 +901,10 @@ static void cgroup_diput(struct dentry *dentry, struct inode *inode) mutex_unlock(&cgroup_mutex); /* - * We want to drop the active superblock reference from the - * cgroup creation after all the dentry refs are gone - - * kill_sb gets mighty unhappy otherwise. Mark - * dentry->d_fsdata with cgroup_diput() to tell - * cgroup_d_release() to call deactivate_super(). + * Drop the active superblock reference that we took when we + * created the cgroup */ - dentry->d_fsdata = cgroup_diput; + deactivate_super(cgrp->root->sb); /* * if we're getting rid of the cgroup, refcount should ensure @@ -933,13 +930,6 @@ static int cgroup_delete(const struct dentry *d) return 1; } -static void cgroup_d_release(struct dentry *dentry) -{ - /* did cgroup_diput() tell me to deactivate super? */ - if (dentry->d_fsdata == cgroup_diput) - deactivate_super(dentry->d_sb); -} - static void remove_dir(struct dentry *d) { struct dentry *parent = dget(d->d_parent); @@ -1547,7 +1537,6 @@ static int cgroup_get_rootdir(struct super_block *sb) static const struct dentry_operations cgroup_dops = { .d_iput = cgroup_diput, .d_delete = cgroup_delete, - .d_release = cgroup_d_release, }; struct inode *inode = -- cgit v0.10.2 From 5db9a4d99b0157a513944e9a44d29c9cec2e91dc Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 7 Jul 2012 16:08:18 -0700 Subject: cgroup: fix cgroup hierarchy umount race 48ddbe1946 "cgroup: make css->refcnt clearing on cgroup removal optional" allowed a css to linger after the associated cgroup is removed. As a css holds a reference on the cgroup's dentry, it means that cgroup dentries may linger for a while. Destroying a superblock which has dentries with positive refcnts is a critical bug and triggers BUG() in vfs code. As each cgroup dentry holds an s_active reference, any lingering cgroup has both its dentry and the superblock pinned and thus preventing premature release of superblock. Unfortunately, after 48ddbe1946, there's a small window while releasing a cgroup which is directly under the root of the hierarchy. When a cgroup directory is released, vfs layer first deletes the corresponding dentry and then invokes dput() on the parent, which may recurse further, so when a cgroup directly below root cgroup is released, the cgroup is first destroyed - which releases the s_active it was holding - and then the dentry for the root cgroup is dput(). This creates a window where the root dentry's refcnt isn't zero but superblock's s_active is. If umount happens before or during this window, vfs will see the root dentry with non-zero refcnt and trigger BUG(). Before 48ddbe1946, this problem didn't exist because the last dentry reference was guaranteed to be put synchronously from rmdir(2) invocation which holds s_active around the whole process. Fix it by holding an extra superblock->s_active reference across dput() from css release, which is the dput() path added by 48ddbe1946 and the only one which doesn't hold an extra s_active ref across the final cgroup dput(). Signed-off-by: Tejun Heo LKML-Reference: <4FEEA5CB.8070809@huawei.com> Reported-by: shyju pv Tested-by: shyju pv Cc: Sasha Levin Acked-by: Li Zefan diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 5f134a0..b303dfc 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -3883,8 +3883,12 @@ static void css_dput_fn(struct work_struct *work) { struct cgroup_subsys_state *css = container_of(work, struct cgroup_subsys_state, dput_work); + struct dentry *dentry = css->cgroup->dentry; + struct super_block *sb = dentry->d_sb; - dput(css->cgroup->dentry); + atomic_inc(&sb->s_active); + dput(dentry); + deactivate_super(sb); } static void init_cgroup_css(struct cgroup_subsys_state *css, -- cgit v0.10.2 From 332a2e1244bd08b9e3ecd378028513396a004a24 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sat, 7 Jul 2012 10:17:00 -0700 Subject: vfs: make O_PATH file descriptors usable for 'fchdir()' MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We already use them for openat() and friends, but fchdir() also wants to be able to use O_PATH file descriptors. This should make it comparable to the O_SEARCH of Solaris. In particular, O_PATH allows you to access (not-quite-open) a directory you don't have read persmission to, only execute permission. Noticed during development of multithread support for ksh93. Reported-by: ольга крыжановская Cc: Al Viro Cc: stable@kernel.org # O_PATH introduced in 3.0+ Signed-off-by: Linus Torvalds diff --git a/fs/open.c b/fs/open.c index d6c79a0..1540632 100644 --- a/fs/open.c +++ b/fs/open.c @@ -397,10 +397,10 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd) { struct file *file; struct inode *inode; - int error; + int error, fput_needed; error = -EBADF; - file = fget(fd); + file = fget_raw_light(fd, &fput_needed); if (!file) goto out; @@ -414,7 +414,7 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd) if (!error) set_fs_pwd(current->fs, &file->f_path); out_putf: - fput(file); + fput_light(file, fput_needed); out: return error; } -- cgit v0.10.2 From bd0a521e88aa7a06ae7aabaed7ae196ed4ad867a Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sat, 7 Jul 2012 17:23:56 -0700 Subject: Linux 3.5-rc6 diff --git a/Makefile b/Makefile index 81ea154..bf02707 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 3 PATCHLEVEL = 5 SUBLEVEL = 0 -EXTRAVERSION = -rc5 +EXTRAVERSION = -rc6 NAME = Saber-toothed Squirrel # *DOCUMENTATION* -- cgit v0.10.2 From 222a806af830fda34ad1f6bc991cd226916de060 Mon Sep 17 00:00:00 2001 From: Mark Rustad Date: Thu, 21 Jun 2012 12:23:42 -0700 Subject: [SCSI] Fix NULL dereferences in scsi_cmd_to_driver Avoid crashing if the private_data pointer happens to be NULL. This has been seen sometimes when a host reset happens, notably when there are many LUNs: host3: Assigned Port ID 0c1601 scsi host3: libfc: Host reset succeeded on port (0c1601) BUG: unable to handle kernel NULL pointer dereference at 0000000000000350 IP: [] scsi_send_eh_cmnd+0x58/0x3a0 Process scsi_eh_3 (pid: 4144, threadinfo ffff88030920c000, task ffff880326b160c0) Stack: 000000010372e6ba 0000000000000282 000027100920dca0 ffffffffa0038ee0 0000000000000000 0000000000030003 ffff88030920dc80 ffff88030920dc80 00000002000e0000 0000000a00004000 ffff8803242f7760 ffff88031326ed80 Call Trace: [] ? lock_timer_base+0x70/0x70 [] scsi_eh_tur+0x3e/0xc0 [] scsi_eh_test_devices+0x76/0x170 [] scsi_eh_host_reset+0x85/0x160 [] scsi_eh_ready_devs+0x91/0x110 [] scsi_unjam_host+0xed/0x1f0 [] scsi_error_handler+0x1a8/0x200 [] ? scsi_unjam_host+0x1f0/0x1f0 [] kthread+0x9e/0xb0 [] kernel_thread_helper+0x4/0x10 [] ? kthread_freezable_should_stop+0x70/0x70 [] ? gs_change+0x13/0x13 Code: 25 28 00 00 00 48 89 45 c8 31 c0 48 8b 87 80 00 00 00 48 8d b5 60 ff ff ff 89 d1 48 89 fb 41 89 d6 4c 89 fa 48 8b 80 b8 00 00 00 <48> 8b 80 50 03 00 00 48 8b 00 48 89 85 38 ff ff ff 48 8b 07 4c RIP [] scsi_send_eh_cmnd+0x58/0x3a0 RSP CR2: 0000000000000350 Signed-off-by: Mark Rustad Tested-by: Marcus Dennis Cc: Signed-off-by: James Bottomley diff --git a/include/scsi/scsi_cmnd.h b/include/scsi/scsi_cmnd.h index 1e11985..ac06cc5 100644 --- a/include/scsi/scsi_cmnd.h +++ b/include/scsi/scsi_cmnd.h @@ -134,10 +134,16 @@ struct scsi_cmnd { static inline struct scsi_driver *scsi_cmd_to_driver(struct scsi_cmnd *cmd) { + struct scsi_driver **sdp; + if (!cmd->request->rq_disk) return NULL; - return *(struct scsi_driver **)cmd->request->rq_disk->private_data; + sdp = (struct scsi_driver **)cmd->request->rq_disk->private_data; + if (!sdp) + return NULL; + + return *sdp; } extern struct scsi_cmnd *scsi_get_command(struct scsi_device *, gfp_t); -- cgit v0.10.2 From 6ef1b512f4e6f936d89aa20be3d97a7ec7c290ac Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Fri, 22 Jun 2012 10:52:34 -0700 Subject: [SCSI] libsas: fix taskfile corruption in sas_ata_qc_fill_rtf fill_result_tf() grabs the taskfile flags from the originating qc which sas_ata_qc_fill_rtf() promptly overwrites. The presence of an ata_taskfile in the sata_device makes it tempting to just copy the full contents in sas_ata_qc_fill_rtf(). However, libata really only wants the fis contents and expects the other portions of the taskfile to not be touched by ->qc_fill_rtf. To that end store a fis buffer in the sata_device and use ata_tf_from_fis() like every other ->qc_fill_rtf() implementation. Cc: Reported-by: Praveen Murali Tested-by: Praveen Murali Signed-off-by: Dan Williams Signed-off-by: James Bottomley diff --git a/drivers/scsi/aic94xx/aic94xx_task.c b/drivers/scsi/aic94xx/aic94xx_task.c index 532d212..393e7ce 100644 --- a/drivers/scsi/aic94xx/aic94xx_task.c +++ b/drivers/scsi/aic94xx/aic94xx_task.c @@ -201,7 +201,7 @@ static void asd_get_response_tasklet(struct asd_ascb *ascb, if (SAS_STATUS_BUF_SIZE >= sizeof(*resp)) { resp->frame_len = le16_to_cpu(*(__le16 *)(r+6)); - memcpy(&resp->ending_fis[0], r+16, 24); + memcpy(&resp->ending_fis[0], r+16, ATA_RESP_FIS_SIZE); ts->buf_valid_size = sizeof(*resp); } } diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c index 441d88a..d109cc3 100644 --- a/drivers/scsi/libsas/sas_ata.c +++ b/drivers/scsi/libsas/sas_ata.c @@ -139,12 +139,12 @@ static void sas_ata_task_done(struct sas_task *task) if (stat->stat == SAS_PROTO_RESPONSE || stat->stat == SAM_STAT_GOOD || ((stat->stat == SAM_STAT_CHECK_CONDITION && dev->sata_dev.command_set == ATAPI_COMMAND_SET))) { - ata_tf_from_fis(resp->ending_fis, &dev->sata_dev.tf); + memcpy(dev->sata_dev.fis, resp->ending_fis, ATA_RESP_FIS_SIZE); if (!link->sactive) { - qc->err_mask |= ac_err_mask(dev->sata_dev.tf.command); + qc->err_mask |= ac_err_mask(dev->sata_dev.fis[2]); } else { - link->eh_info.err_mask |= ac_err_mask(dev->sata_dev.tf.command); + link->eh_info.err_mask |= ac_err_mask(dev->sata_dev.fis[2]); if (unlikely(link->eh_info.err_mask)) qc->flags |= ATA_QCFLAG_FAILED; } @@ -161,8 +161,8 @@ static void sas_ata_task_done(struct sas_task *task) qc->flags |= ATA_QCFLAG_FAILED; } - dev->sata_dev.tf.feature = 0x04; /* status err */ - dev->sata_dev.tf.command = ATA_ERR; + dev->sata_dev.fis[3] = 0x04; /* status err */ + dev->sata_dev.fis[2] = ATA_ERR; } } @@ -269,7 +269,7 @@ static bool sas_ata_qc_fill_rtf(struct ata_queued_cmd *qc) { struct domain_device *dev = qc->ap->private_data; - memcpy(&qc->result_tf, &dev->sata_dev.tf, sizeof(qc->result_tf)); + ata_tf_from_fis(dev->sata_dev.fis, &qc->result_tf); return true; } diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h index f4f1c96..10ce74f 100644 --- a/include/scsi/libsas.h +++ b/include/scsi/libsas.h @@ -163,6 +163,8 @@ enum ata_command_set { ATAPI_COMMAND_SET = 1, }; +#define ATA_RESP_FIS_SIZE 24 + struct sata_device { enum ata_command_set command_set; struct smp_resp rps_resp; /* report_phy_sata_resp */ @@ -171,7 +173,7 @@ struct sata_device { struct ata_port *ap; struct ata_host ata_host; - struct ata_taskfile tf; + u8 fis[ATA_RESP_FIS_SIZE]; }; enum { @@ -537,7 +539,7 @@ enum exec_status { */ struct ata_task_resp { u16 frame_len; - u8 ending_fis[24]; /* dev to host or data-in */ + u8 ending_fis[ATA_RESP_FIS_SIZE]; /* dev to host or data-in */ }; #define SAS_STATUS_BUF_SIZE 96 -- cgit v0.10.2 From a77171806515fb5e2288219ddb47af1f0b1328e7 Mon Sep 17 00:00:00 2001 From: Eddie Wai Date: Fri, 29 Jun 2012 16:37:35 -0700 Subject: [SCSI] bnx2i: Removed the reference to the netdev->base_addr The netdev->base_addr parameter has been deprecated in the L2 bnx2 driver. This is used by bnx2i for the BARn iomapping. This patch will directly reference the pci_resource_start instead of using the deprecated netdev->base_addr. This patch is actually a critical bug fix as the 1G bnx2 driver no longer supports the netdev->base_addr in the current kernel of the scsi tree. This means that Broadcom's 1G Linux iSCSI offload solution would not work at all without this patch. Signed-off-by: Eddie Wai Reviewed-by: Mike Christie Signed-off-by: James Bottomley diff --git a/drivers/scsi/bnx2i/bnx2i.h b/drivers/scsi/bnx2i/bnx2i.h index 0c53c28..7e77cf6 100644 --- a/drivers/scsi/bnx2i/bnx2i.h +++ b/drivers/scsi/bnx2i/bnx2i.h @@ -350,6 +350,7 @@ struct bnx2i_hba { struct pci_dev *pcidev; struct net_device *netdev; void __iomem *regview; + resource_size_t reg_base; u32 age; unsigned long cnic_dev_type; diff --git a/drivers/scsi/bnx2i/bnx2i_hwi.c b/drivers/scsi/bnx2i/bnx2i_hwi.c index ece47e5..86a12b4 100644 --- a/drivers/scsi/bnx2i/bnx2i_hwi.c +++ b/drivers/scsi/bnx2i/bnx2i_hwi.c @@ -2724,7 +2724,6 @@ int bnx2i_map_ep_dbell_regs(struct bnx2i_endpoint *ep) goto arm_cq; } - reg_base = ep->hba->netdev->base_addr; if ((test_bit(BNX2I_NX2_DEV_5709, &ep->hba->cnic_dev_type)) && (ep->hba->mail_queue_access == BNX2I_MQ_BIN_MODE)) { config2 = REG_RD(ep->hba, BNX2_MQ_CONFIG2); @@ -2740,7 +2739,7 @@ int bnx2i_map_ep_dbell_regs(struct bnx2i_endpoint *ep) /* 5709 device in normal node and 5706/5708 devices */ reg_off = CTX_OFFSET + (MB_KERNEL_CTX_SIZE * cid_num); - ep->qp.ctx_base = ioremap_nocache(reg_base + reg_off, + ep->qp.ctx_base = ioremap_nocache(ep->hba->reg_base + reg_off, MB_KERNEL_CTX_SIZE); if (!ep->qp.ctx_base) return -ENOMEM; diff --git a/drivers/scsi/bnx2i/bnx2i_iscsi.c b/drivers/scsi/bnx2i/bnx2i_iscsi.c index f8d516b..621538b 100644 --- a/drivers/scsi/bnx2i/bnx2i_iscsi.c +++ b/drivers/scsi/bnx2i/bnx2i_iscsi.c @@ -811,13 +811,13 @@ struct bnx2i_hba *bnx2i_alloc_hba(struct cnic_dev *cnic) bnx2i_identify_device(hba); bnx2i_setup_host_queue_size(hba, shost); + hba->reg_base = pci_resource_start(hba->pcidev, 0); if (test_bit(BNX2I_NX2_DEV_5709, &hba->cnic_dev_type)) { - hba->regview = ioremap_nocache(hba->netdev->base_addr, - BNX2_MQ_CONFIG2); + hba->regview = pci_iomap(hba->pcidev, 0, BNX2_MQ_CONFIG2); if (!hba->regview) goto ioreg_map_err; } else if (test_bit(BNX2I_NX2_DEV_57710, &hba->cnic_dev_type)) { - hba->regview = ioremap_nocache(hba->netdev->base_addr, 4096); + hba->regview = pci_iomap(hba->pcidev, 0, 4096); if (!hba->regview) goto ioreg_map_err; } @@ -884,7 +884,7 @@ cid_que_err: bnx2i_free_mp_bdt(hba); mp_bdt_mem_err: if (hba->regview) { - iounmap(hba->regview); + pci_iounmap(hba->pcidev, hba->regview); hba->regview = NULL; } ioreg_map_err: @@ -910,7 +910,7 @@ void bnx2i_free_hba(struct bnx2i_hba *hba) pci_dev_put(hba->pcidev); if (hba->regview) { - iounmap(hba->regview); + pci_iounmap(hba->pcidev, hba->regview); hba->regview = NULL; } bnx2i_free_mp_bdt(hba); -- cgit v0.10.2 From 736f29cd6b7af9646a21a34866cd16277e03ee69 Mon Sep 17 00:00:00 2001 From: Tomi Valkeinen Date: Wed, 4 Jul 2012 18:13:48 +0530 Subject: OMAPDSS: Use PM notifiers for system suspend The current way how omapdss handles system suspend and resume is that omapdss device (a platform device, which is not part of the device hierarchy of the DSS HW devices, like DISPC and DSI, or panels.) uses the suspend and resume callbacks from platform_driver to handle system suspend. It does this by disabling all enabled panels on suspend, and resuming the previously disabled panels on resume. This presents a few problems. One is that as omapdss device is not related to the panel devices or the DSS HW devices, there's no ordering in the suspend process. This means that suspend could be first ran for DSS HW devices and panels, and only then for omapdss device. Currently this is not a problem, as DSS HW devices and panels do not handle suspend. Another, more pressing problem, is that when suspending or resuming, the runtime PM functions return -EACCES as runtime PM is disabled during system suspend. This causes the driver to print warnings, and operations to fail as they think that they failed to bring up the HW. This patch changes the omapdss suspend handling to use PM notifiers, which are called before suspend and after resume. This way we have a normally functioning system when we are suspending and resuming the panels. This patch, I believe, creates a problem that somebody could enable or disable a panel between PM_SUSPEND_PREPARE and the system suspend, and similarly the other way around in resume. I choose to ignore the problem for now, as it sounds rather unlikely, and if it happens, it's not fatal. In the long run the system suspend handling of omapdss and panels should be thought out properly. The current approach feels rather hacky. Perhaps the panel drivers should handle system suspend, or the users of omapdss (omapfb, omapdrm) should handle system suspend. Note that after this patch we could probably revert 0eaf9f52e94f756147dbfe1faf1f77a02378dbf9 (OMAPDSS: use sync versions of pm_runtime_put). But as I said, this patch may be temporary, so let's leave the sync version still in place. Signed-off-by: Tomi Valkeinen Reported-by: Jassi Brar Tested-by: Jassi Brar Tested-by: Joe Woodward Signed-off-by: Archit Taneja [fts: fixed 2 brace coding style issues] Signed-off-by: Florian Tobias Schandinat diff --git a/drivers/video/omap2/dss/core.c b/drivers/video/omap2/dss/core.c index 5066eee..58bd9c2 100644 --- a/drivers/video/omap2/dss/core.c +++ b/drivers/video/omap2/dss/core.c @@ -32,6 +32,7 @@ #include #include #include +#include #include