From 2aba76f014a7b56ab4fe75845c5fd57b5590acc2 Mon Sep 17 00:00:00 2001 From: Michael Williamson Date: Fri, 20 May 2011 10:26:06 -0400 Subject: audio: tlv320aic26: fix PLL register configuration The current PLL configuration code for the tlc320aic26 codec appears to assume a hardcoded system clock of 12 MHz. Use the clock value provided by the DAI_OPS API for the calculation. Tested using a MityDSP-L138 platform providing a 24.576 MHz clock. Signed-off-by: Michael Williamson Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/sound/soc/codecs/tlv320aic26.c b/sound/soc/codecs/tlv320aic26.c index e2a7608..7859bdc 100644 --- a/sound/soc/codecs/tlv320aic26.c +++ b/sound/soc/codecs/tlv320aic26.c @@ -161,10 +161,18 @@ static int aic26_hw_params(struct snd_pcm_substream *substream, dev_dbg(&aic26->spi->dev, "bad format\n"); return -EINVAL; } - /* Configure PLL */ + /** + * Configure PLL + * fsref = (mclk * PLLM) / 2048 + * where PLLM = J.DDDD (DDDD register ranges from 0 to 9999, decimal) + */ pval = 1; - jval = (fsref == 44100) ? 7 : 8; - dval = (fsref == 44100) ? 5264 : 1920; + /* compute J portion of multiplier */ + jval = fsref / (aic26->mclk / 2048); + /* compute fractional DDDD component of multiplier */ + dval = fsref - (jval * (aic26->mclk / 2048)); + dval = (10000 * dval) / (aic26->mclk / 2048); + dev_dbg(&aic26->spi->dev, "Setting PLLM to %d.%04d\n", jval, dval); qval = 0; reg = 0x8000 | qval << 11 | pval << 8 | jval << 2; aic26_reg_write(codec, AIC26_REG_PLL_PROG1, reg); -- cgit v0.10.2 From 508b76864c18f34f8d6ba08d192f5817f8dc8ead Mon Sep 17 00:00:00 2001 From: Jarkko Nikula Date: Fri, 20 May 2011 16:52:37 +0300 Subject: ASoC: tlv320aic3x: Don't sync first two registers from register cache There is no need to sync first two registers from cache to hw after a reset. First one is used to select page for register access and this driver is normally accessing page 0 only. Second one does a software reset which is obviously unneeded after hardware or previous software reset command. Signed-off-by: Jarkko Nikula Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/sound/soc/codecs/tlv320aic3x.c b/sound/soc/codecs/tlv320aic3x.c index c3d96fc..9047bb1 100644 --- a/sound/soc/codecs/tlv320aic3x.c +++ b/sound/soc/codecs/tlv320aic3x.c @@ -1114,7 +1114,7 @@ static int aic3x_set_power(struct snd_soc_codec *codec, int power) /* Sync reg_cache with the hardware */ codec->cache_only = 0; - for (i = 0; i < ARRAY_SIZE(aic3x_reg); i++) + for (i = AIC3X_SAMPLE_RATE_SEL_REG; i < ARRAY_SIZE(aic3x_reg); i++) snd_soc_write(codec, i, cache[i]); if (aic3x->model == AIC3X_MODEL_3007) aic3x_init_3007(codec); -- cgit v0.10.2 From 9fb352b18b11124ed1ddebc0d74ebbd7ba8defd7 Mon Sep 17 00:00:00 2001 From: Jarkko Nikula Date: Fri, 20 May 2011 16:52:38 +0300 Subject: ASoC: tlv320aic3x: Do soft reset to codec when going to bias off state TLV320AIC33, TLV320AIC34 and I believe others too in this family have some hw bugs that cause that analogue and digital VDD supplies remain leaking up to a few mA of current after certain use cases even the hw blocks inside codec are driven to off. Highest leakages occur after using the bypass paths inside codec but it is possible to get smaller leakages just by toggling mute switches in unused audio paths (i.e. no DAPM changes) while codec is on due another active audio path. While some cases are able to workaroud by making sure that e.g. output mixer switches are muted before powering down the output stage this doesn't help all the cases. Therefore use the software reset command to clear possible leakage currents since that works in every cases and affects only this codec instance. Only drawback is that now cache sync is required everytime when codec bias comes out from bias off state, not only when supply regulators were off. Signed-off-by: Jarkko Nikula Acked-by: Mark Brown Signed-off-by: Liam Girdwood diff --git a/sound/soc/codecs/tlv320aic3x.c b/sound/soc/codecs/tlv320aic3x.c index 9047bb1..789453d 100644 --- a/sound/soc/codecs/tlv320aic3x.c +++ b/sound/soc/codecs/tlv320aic3x.c @@ -1120,6 +1120,13 @@ static int aic3x_set_power(struct snd_soc_codec *codec, int power) aic3x_init_3007(codec); codec->cache_sync = 0; } else { + /* + * Do soft reset to this codec instance in order to clear + * possible VDD leakage currents in case the supply regulators + * remain on + */ + snd_soc_write(codec, AIC3X_RESET, SOFT_RESET); + codec->cache_sync = 1; aic3x->power = 0; /* HW writes are needless when bias is off */ codec->cache_only = 1; -- cgit v0.10.2 From 7f77897ef2b6a5ee4eb8bc24fe8b1f3eab254328 Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Tue, 24 May 2011 11:43:18 +0200 Subject: HID: hiddev: fix potential use-after-free Commit 6cb4b040795 ("HID: hiddev: fix race between hiddev_disconnect and hiddev_release") made it possible to access hiddev (for unlocking the existance mutex) once hiddev has been kfreed. Change the order so that this can not happen (always unlock the mutex first, it is needed only to protect access to ->exist and ->open). Signed-off-by: Jiri Kosina diff --git a/drivers/hid/usbhid/hiddev.c b/drivers/hid/usbhid/hiddev.c index ff3c644..4985f485 100644 --- a/drivers/hid/usbhid/hiddev.c +++ b/drivers/hid/usbhid/hiddev.c @@ -923,10 +923,11 @@ void hiddev_disconnect(struct hid_device *hid) usb_deregister_dev(usbhid->intf, &hiddev_class); if (hiddev->open) { + mutex_unlock(&hiddev->existancelock); usbhid_close(hiddev->hid); wake_up_interruptible(&hiddev->wait); } else { + mutex_unlock(&hiddev->existancelock); kfree(hiddev); } - mutex_unlock(&hiddev->existancelock); } -- cgit v0.10.2 From e23be0a27dcc9297ff0495360d89bc5b0bf12383 Mon Sep 17 00:00:00 2001 From: Jimmy Hon Date: Fri, 20 May 2011 17:59:19 -0400 Subject: HID: add quirk for HyperPen 10000U Add 5543:0064 UC-Logic Technology Corp. Aiptek HyperPen 10000U to quirks with HID_QUIRK_MULTI_INPUT. Originally the device is reporting the x,y coordinates on Z and RX. By adding this quirk, there will be two kernel devices. The first one is muted and the second device will report coordinates on X and Y. Signed-off-by: Jimmy Hon Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 0b374a6..2bbaad7 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -622,6 +622,7 @@ #define USB_VENDOR_ID_UCLOGIC 0x5543 #define USB_DEVICE_ID_UCLOGIC_TABLET_PF1209 0x0042 #define USB_DEVICE_ID_UCLOGIC_TABLET_KNA5 0x6001 +#define USB_DEVICE_ID_UCLOGIC_TABLET_TWA60 0x0064 #define USB_DEVICE_ID_UCLOGIC_TABLET_WP4030U 0x0003 #define USB_DEVICE_ID_UCLOGIC_TABLET_WP5540U 0x0004 #define USB_DEVICE_ID_UCLOGIC_TABLET_WP8060U 0x0005 diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c index 0e30b14..621959d 100644 --- a/drivers/hid/usbhid/hid-quirks.c +++ b/drivers/hid/usbhid/hid-quirks.c @@ -74,6 +74,7 @@ static const struct hid_blacklist { { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_PF1209, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_WP4030U, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_KNA5, HID_QUIRK_MULTI_INPUT }, + { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_TWA60, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_WP5540U, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_UCLOGIC, USB_DEVICE_ID_UCLOGIC_TABLET_WP8060U, HID_QUIRK_MULTI_INPUT }, { USB_VENDOR_ID_WALTOP, USB_DEVICE_ID_WALTOP_MEDIA_TABLET_10_6_INCH, HID_QUIRK_MULTI_INPUT }, -- cgit v0.10.2 From ac10f81d94f49f1bd9618680263400d275ddf825 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 25 May 2011 01:00:45 -0400 Subject: drm/radeon/kms/blit: workaround some hw issues on evergreen+ Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen_blit_kms.c b/drivers/gpu/drm/radeon/evergreen_blit_kms.c index ba06a69..4086729 100644 --- a/drivers/gpu/drm/radeon/evergreen_blit_kms.c +++ b/drivers/gpu/drm/radeon/evergreen_blit_kms.c @@ -199,6 +199,16 @@ static void set_scissors(struct radeon_device *rdev, int x1, int y1, int x2, int y2) { + /* workaround some hw bugs */ + if (x2 == 0) + x1 = 1; + if (y2 == 0) + y1 = 1; + if (rdev->family == CHIP_CAYMAN) { + if ((x2 == 1) && (y2 == 1)) + x2 = 2; + } + radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); radeon_ring_write(rdev, (PA_SC_SCREEN_SCISSOR_TL - PACKET3_SET_CONTEXT_REG_START) >> 2); radeon_ring_write(rdev, (x1 << 0) | (y1 << 16)); -- cgit v0.10.2 From f90e5b5b136ede1f0fd15999e95f13124d6b0dbd Mon Sep 17 00:00:00 2001 From: Bob Peterson Date: Tue, 24 May 2011 10:44:42 -0400 Subject: GFS2: Processes waiting on inode glock that no processes are holding This patch fixes a race in the GFS2 glock state machine that may result in lockups. The symptom is that all nodes but one will hang, waiting for a particular glock. All the holder records will have the "W" (Waiting) bit set. The other node will typically have the glock stuck in Exclusive mode (EX) with no holder records, but the dinode will be cached. In other words, an entry with "I:" will appear in the glock dump for that glock, but nothing else. The race has to do with the glock "Pending Demote" bit, which can be set, then immediately reset, thus losing the fact that another node needs the glock. The sequence of events is: 1. Something schedules the glock workqueue (e.g. glock request from fs) 2. The glock workqueue gets to the point between the test of the reply pending bit and the spin lock: if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags)) { finish_xmote(gl, gl->gl_reply); drop_ref = 1; } down_read(&gfs2_umount_flush_sem); <---- i.e. here spin_lock(&gl->gl_spin); 3. In comes (a) the reply to our EX lock request setting GLF_REPLY_PENDING and (b) the demote request which sets GLF_PENDING_DEMOTE 4. The following test is executed: if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) && gl->gl_state != LM_ST_UNLOCKED && gl->gl_demote_state != LM_ST_EXCLUSIVE) { This resets the pending demote flag, and gl->gl_demote_state is not equal to exclusive, however because the reply from the dlm arrived after we checked for the GLF_REPLY_PENDING flag, gl->gl_state is still equal to unlocked, so although we reset the GLF_PENDING_DEMOTE flag, we didn't then set the GLF_DEMOTE flag or reinstate the GLF_PENDING_DEMOTE_FLAG. The patch closes the timing window by only transitioning the "Pending demote" bit to the "demote" flag once we know the other conditions (not unlocked and not exclusive) are met. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index a2a6abb..7137750 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -663,14 +663,19 @@ static void glock_work_func(struct work_struct *work) drop_ref = 1; } spin_lock(&gl->gl_spin); - if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) && + if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) && gl->gl_state != LM_ST_UNLOCKED && gl->gl_demote_state != LM_ST_EXCLUSIVE) { unsigned long holdtime, now = jiffies; + holdtime = gl->gl_tchange + gl->gl_ops->go_min_hold_time; if (time_before(now, holdtime)) delay = holdtime - now; - set_bit(delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE, &gl->gl_flags); + + if (!delay) { + clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags); + set_bit(GLF_DEMOTE, &gl->gl_flags); + } } run_queue(gl, 0); spin_unlock(&gl->gl_spin); -- cgit v0.10.2 From 8c127f0717b438e6abc3d92d4ae248c4224b9dcb Mon Sep 17 00:00:00 2001 From: Hans Petter Selasky Date: Wed, 25 May 2011 09:24:32 -0700 Subject: Input: properly assign return value of clamp() macro. [dtor@mail.ru: added mousedev changes] Signed-off-by: Hans Petter Selasky Cc: stable@kernel.org Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/input.c b/drivers/input/input.c index 75e11c7..da38d97 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -1756,7 +1756,7 @@ static unsigned int input_estimate_events_per_packet(struct input_dev *dev) } else if (test_bit(ABS_MT_TRACKING_ID, dev->absbit)) { mt_slots = dev->absinfo[ABS_MT_TRACKING_ID].maximum - dev->absinfo[ABS_MT_TRACKING_ID].minimum + 1, - clamp(mt_slots, 2, 32); + mt_slots = clamp(mt_slots, 2, 32); } else if (test_bit(ABS_MT_POSITION_X, dev->absbit)) { mt_slots = 2; } else { diff --git a/drivers/input/mousedev.c b/drivers/input/mousedev.c index 257e033..0110b5a 100644 --- a/drivers/input/mousedev.c +++ b/drivers/input/mousedev.c @@ -187,7 +187,7 @@ static void mousedev_abs_event(struct input_dev *dev, struct mousedev *mousedev, if (size == 0) size = xres ? : 1; - clamp(value, min, max); + value = clamp(value, min, max); mousedev->packet.x = ((value - min) * xres) / size; mousedev->packet.abs_event = 1; @@ -201,7 +201,7 @@ static void mousedev_abs_event(struct input_dev *dev, struct mousedev *mousedev, if (size == 0) size = yres ? : 1; - clamp(value, min, max); + value = clamp(value, min, max); mousedev->packet.y = yres - ((value - min) * yres) / size; mousedev->packet.abs_event = 1; -- cgit v0.10.2 From cb92d452ba665205ad6bfb424c0ef009cf26587d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 25 May 2011 16:39:00 -0400 Subject: drm/radeon/kms: add blit support for cayman (v2) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Allows us to use the 3D engine for memory management and allows us to use vram beyond the BAR aperture. v2: fix copy paste typo Reported-by: Nils Wallménius Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/cayman_blit_shaders.c b/drivers/gpu/drm/radeon/cayman_blit_shaders.c index e148ab0..7b4eeb7 100644 --- a/drivers/gpu/drm/radeon/cayman_blit_shaders.c +++ b/drivers/gpu/drm/radeon/cayman_blit_shaders.c @@ -39,17 +39,335 @@ const u32 cayman_default_state[] = { - /* XXX fill in additional blit state */ + 0xc0066900, + 0x00000000, + 0x00000060, /* DB_RENDER_CONTROL */ + 0x00000000, /* DB_COUNT_CONTROL */ + 0x00000000, /* DB_DEPTH_VIEW */ + 0x0000002a, /* DB_RENDER_OVERRIDE */ + 0x00000000, /* DB_RENDER_OVERRIDE2 */ + 0x00000000, /* DB_HTILE_DATA_BASE */ 0xc0026900, - 0x00000316, - 0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */ - 0x00000010, /* */ + 0x0000000a, + 0x00000000, /* DB_STENCIL_CLEAR */ + 0x00000000, /* DB_DEPTH_CLEAR */ + + 0xc0036900, + 0x0000000f, + 0x00000000, /* DB_DEPTH_INFO */ + 0x00000000, /* DB_Z_INFO */ + 0x00000000, /* DB_STENCIL_INFO */ + + 0xc0016900, + 0x00000080, + 0x00000000, /* PA_SC_WINDOW_OFFSET */ + + 0xc00d6900, + 0x00000083, + 0x0000ffff, /* PA_SC_CLIPRECT_RULE */ + 0x00000000, /* PA_SC_CLIPRECT_0_TL */ + 0x20002000, /* PA_SC_CLIPRECT_0_BR */ + 0x00000000, + 0x20002000, + 0x00000000, + 0x20002000, + 0x00000000, + 0x20002000, + 0xaaaaaaaa, /* PA_SC_EDGERULE */ + 0x00000000, /* PA_SU_HARDWARE_SCREEN_OFFSET */ + 0x0000000f, /* CB_TARGET_MASK */ + 0x0000000f, /* CB_SHADER_MASK */ + + 0xc0226900, + 0x00000094, + 0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */ + 0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */ + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x00000000, /* PA_SC_VPORT_ZMIN_0 */ + 0x3f800000, /* PA_SC_VPORT_ZMAX_0 */ + + 0xc0016900, + 0x000000d4, + 0x00000000, /* SX_MISC */ 0xc0026900, 0x000000d9, 0x00000000, /* CP_RINGID */ 0x00000000, /* CP_VMID */ + + 0xc0096900, + 0x00000100, + 0x00ffffff, /* VGT_MAX_VTX_INDX */ + 0x00000000, /* VGT_MIN_VTX_INDX */ + 0x00000000, /* VGT_INDX_OFFSET */ + 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */ + 0x00000000, /* SX_ALPHA_TEST_CONTROL */ + 0x00000000, /* CB_BLEND_RED */ + 0x00000000, /* CB_BLEND_GREEN */ + 0x00000000, /* CB_BLEND_BLUE */ + 0x00000000, /* CB_BLEND_ALPHA */ + + 0xc0016900, + 0x00000187, + 0x00000100, /* SPI_VS_OUT_ID_0 */ + + 0xc0026900, + 0x00000191, + 0x00000100, /* SPI_PS_INPUT_CNTL_0 */ + 0x00000101, /* SPI_PS_INPUT_CNTL_1 */ + + 0xc0016900, + 0x000001b1, + 0x00000000, /* SPI_VS_OUT_CONFIG */ + + 0xc0106900, + 0x000001b3, + 0x20000001, /* SPI_PS_IN_CONTROL_0 */ + 0x00000000, /* SPI_PS_IN_CONTROL_1 */ + 0x00000000, /* SPI_INTERP_CONTROL_0 */ + 0x00000000, /* SPI_INPUT_Z */ + 0x00000000, /* SPI_FOG_CNTL */ + 0x00100000, /* SPI_BARYC_CNTL */ + 0x00000000, /* SPI_PS_IN_CONTROL_2 */ + 0x00000000, /* SPI_COMPUTE_INPUT_CNTL */ + 0x00000000, /* SPI_COMPUTE_NUM_THREAD_X */ + 0x00000000, /* SPI_COMPUTE_NUM_THREAD_Y */ + 0x00000000, /* SPI_COMPUTE_NUM_THREAD_Z */ + 0x00000000, /* SPI_GPR_MGMT */ + 0x00000000, /* SPI_LDS_MGMT */ + 0x00000000, /* SPI_STACK_MGMT */ + 0x00000000, /* SPI_WAVE_MGMT_1 */ + 0x00000000, /* SPI_WAVE_MGMT_2 */ + + 0xc0016900, + 0x000001e0, + 0x00000000, /* CB_BLEND0_CONTROL */ + + 0xc00e6900, + 0x00000200, + 0x00000000, /* DB_DEPTH_CONTROL */ + 0x00000000, /* DB_EQAA */ + 0x00cc0010, /* CB_COLOR_CONTROL */ + 0x00000210, /* DB_SHADER_CONTROL */ + 0x00010000, /* PA_CL_CLIP_CNTL */ + 0x00000004, /* PA_SU_SC_MODE_CNTL */ + 0x00000100, /* PA_CL_VTE_CNTL */ + 0x00000000, /* PA_CL_VS_OUT_CNTL */ + 0x00000000, /* PA_CL_NANINF_CNTL */ + 0x00000000, /* PA_SU_LINE_STIPPLE_CNTL */ + 0x00000000, /* PA_SU_LINE_STIPPLE_SCALE */ + 0x00000000, /* PA_SU_PRIM_FILTER_CNTL */ + 0x00000000, /* */ + 0x00000000, /* */ + + 0xc0026900, + 0x00000229, + 0x00000000, /* SQ_PGM_START_FS */ + 0x00000000, + + 0xc0016900, + 0x0000023b, + 0x00000000, /* SQ_LDS_ALLOC_PS */ + + 0xc0066900, + 0x00000240, + 0x00000000, /* SQ_ESGS_RING_ITEMSIZE */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + + 0xc0046900, + 0x00000247, + 0x00000000, /* SQ_GS_VERT_ITEMSIZE */ + 0x00000000, + 0x00000000, + 0x00000000, + + 0xc0116900, + 0x00000280, + 0x00000000, /* PA_SU_POINT_SIZE */ + 0x00000000, /* PA_SU_POINT_MINMAX */ + 0x00000008, /* PA_SU_LINE_CNTL */ + 0x00000000, /* PA_SC_LINE_STIPPLE */ + 0x00000000, /* VGT_OUTPUT_PATH_CNTL */ + 0x00000000, /* VGT_HOS_CNTL */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, /* VGT_GS_MODE */ + + 0xc0026900, + 0x00000292, + 0x00000000, /* PA_SC_MODE_CNTL_0 */ + 0x00000000, /* PA_SC_MODE_CNTL_1 */ + + 0xc0016900, + 0x000002a1, + 0x00000000, /* VGT_PRIMITIVEID_EN */ + + 0xc0016900, + 0x000002a5, + 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_EN */ + + 0xc0026900, + 0x000002a8, + 0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */ + 0x00000000, + + 0xc0026900, + 0x000002ad, + 0x00000000, /* VGT_REUSE_OFF */ + 0x00000000, + + 0xc0016900, + 0x000002d5, + 0x00000000, /* VGT_SHADER_STAGES_EN */ + + 0xc0016900, + 0x000002dc, + 0x0000aa00, /* DB_ALPHA_TO_MASK */ + + 0xc0066900, + 0x000002de, + 0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + + 0xc0026900, + 0x000002e5, + 0x00000000, /* VGT_STRMOUT_CONFIG */ + 0x00000000, + + 0xc01b6900, + 0x000002f5, + 0x76543210, /* PA_SC_CENTROID_PRIORITY_0 */ + 0xfedcba98, /* PA_SC_CENTROID_PRIORITY_1 */ + 0x00000000, /* PA_SC_LINE_CNTL */ + 0x00000000, /* PA_SC_AA_CONFIG */ + 0x00000005, /* PA_SU_VTX_CNTL */ + 0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */ + 0x3f800000, /* PA_CL_GB_VERT_DISC_ADJ */ + 0x3f800000, /* PA_CL_GB_HORZ_CLIP_ADJ */ + 0x3f800000, /* PA_CL_GB_HORZ_DISC_ADJ */ + 0x00000000, /* PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0 */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0xffffffff, /* PA_SC_AA_MASK_X0Y0_X1Y0 */ + 0xffffffff, + + 0xc0026900, + 0x00000316, + 0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */ + 0x00000010, /* */ +}; + +const u32 cayman_vs[] = +{ + 0x00000004, + 0x80400400, + 0x0000a03c, + 0x95000688, + 0x00004000, + 0x15000688, + 0x00000000, + 0x88000000, + 0x04000000, + 0x67961001, +#ifdef __BIG_ENDIAN + 0x00020000, +#else + 0x00000000, +#endif + 0x00000000, + 0x04000000, + 0x67961000, +#ifdef __BIG_ENDIAN + 0x00020008, +#else + 0x00000008, +#endif + 0x00000000, +}; + +const u32 cayman_ps[] = +{ + 0x00000004, + 0xa00c0000, + 0x00000008, + 0x80400000, + 0x00000000, + 0x95000688, + 0x00000000, + 0x88000000, + 0x00380400, + 0x00146b10, + 0x00380000, + 0x20146b10, + 0x00380400, + 0x40146b00, + 0x80380000, + 0x60146b00, + 0x00000010, + 0x000d1000, + 0xb0800000, + 0x00000000, }; +const u32 cayman_ps_size = ARRAY_SIZE(cayman_ps); +const u32 cayman_vs_size = ARRAY_SIZE(cayman_vs); const u32 cayman_default_size = ARRAY_SIZE(cayman_default_state); diff --git a/drivers/gpu/drm/radeon/cayman_blit_shaders.h b/drivers/gpu/drm/radeon/cayman_blit_shaders.h index 33b75e5d..f5d0e9a 100644 --- a/drivers/gpu/drm/radeon/cayman_blit_shaders.h +++ b/drivers/gpu/drm/radeon/cayman_blit_shaders.h @@ -25,8 +25,11 @@ #ifndef CAYMAN_BLIT_SHADERS_H #define CAYMAN_BLIT_SHADERS_H +extern const u32 cayman_ps[]; +extern const u32 cayman_vs[]; extern const u32 cayman_default_state[]; +extern const u32 cayman_ps_size, cayman_vs_size; extern const u32 cayman_default_size; #endif diff --git a/drivers/gpu/drm/radeon/evergreen_blit_kms.c b/drivers/gpu/drm/radeon/evergreen_blit_kms.c index 4086729..a60ad28 100644 --- a/drivers/gpu/drm/radeon/evergreen_blit_kms.c +++ b/drivers/gpu/drm/radeon/evergreen_blit_kms.c @@ -31,6 +31,7 @@ #include "evergreend.h" #include "evergreen_blit_shaders.h" +#include "cayman_blit_shaders.h" #define DI_PT_RECTLIST 0x11 #define DI_INDEX_SIZE_16_BIT 0x0 @@ -265,238 +266,240 @@ set_default_state(struct radeon_device *rdev) u64 gpu_addr; int dwords; - switch (rdev->family) { - case CHIP_CEDAR: - default: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 96; - num_vs_threads = 16; - num_gs_threads = 16; - num_es_threads = 16; - num_hs_threads = 16; - num_ls_threads = 16; - num_ps_stack_entries = 42; - num_vs_stack_entries = 42; - num_gs_stack_entries = 42; - num_es_stack_entries = 42; - num_hs_stack_entries = 42; - num_ls_stack_entries = 42; - break; - case CHIP_REDWOOD: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 128; - num_vs_threads = 20; - num_gs_threads = 20; - num_es_threads = 20; - num_hs_threads = 20; - num_ls_threads = 20; - num_ps_stack_entries = 42; - num_vs_stack_entries = 42; - num_gs_stack_entries = 42; - num_es_stack_entries = 42; - num_hs_stack_entries = 42; - num_ls_stack_entries = 42; - break; - case CHIP_JUNIPER: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 128; - num_vs_threads = 20; - num_gs_threads = 20; - num_es_threads = 20; - num_hs_threads = 20; - num_ls_threads = 20; - num_ps_stack_entries = 85; - num_vs_stack_entries = 85; - num_gs_stack_entries = 85; - num_es_stack_entries = 85; - num_hs_stack_entries = 85; - num_ls_stack_entries = 85; - break; - case CHIP_CYPRESS: - case CHIP_HEMLOCK: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 128; - num_vs_threads = 20; - num_gs_threads = 20; - num_es_threads = 20; - num_hs_threads = 20; - num_ls_threads = 20; - num_ps_stack_entries = 85; - num_vs_stack_entries = 85; - num_gs_stack_entries = 85; - num_es_stack_entries = 85; - num_hs_stack_entries = 85; - num_ls_stack_entries = 85; - break; - case CHIP_PALM: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 96; - num_vs_threads = 16; - num_gs_threads = 16; - num_es_threads = 16; - num_hs_threads = 16; - num_ls_threads = 16; - num_ps_stack_entries = 42; - num_vs_stack_entries = 42; - num_gs_stack_entries = 42; - num_es_stack_entries = 42; - num_hs_stack_entries = 42; - num_ls_stack_entries = 42; - break; - case CHIP_BARTS: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 128; - num_vs_threads = 20; - num_gs_threads = 20; - num_es_threads = 20; - num_hs_threads = 20; - num_ls_threads = 20; - num_ps_stack_entries = 85; - num_vs_stack_entries = 85; - num_gs_stack_entries = 85; - num_es_stack_entries = 85; - num_hs_stack_entries = 85; - num_ls_stack_entries = 85; - break; - case CHIP_TURKS: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 128; - num_vs_threads = 20; - num_gs_threads = 20; - num_es_threads = 20; - num_hs_threads = 20; - num_ls_threads = 20; - num_ps_stack_entries = 42; - num_vs_stack_entries = 42; - num_gs_stack_entries = 42; - num_es_stack_entries = 42; - num_hs_stack_entries = 42; - num_ls_stack_entries = 42; - break; - case CHIP_CAICOS: - num_ps_gprs = 93; - num_vs_gprs = 46; - num_temp_gprs = 4; - num_gs_gprs = 31; - num_es_gprs = 31; - num_hs_gprs = 23; - num_ls_gprs = 23; - num_ps_threads = 128; - num_vs_threads = 10; - num_gs_threads = 10; - num_es_threads = 10; - num_hs_threads = 10; - num_ls_threads = 10; - num_ps_stack_entries = 42; - num_vs_stack_entries = 42; - num_gs_stack_entries = 42; - num_es_stack_entries = 42; - num_hs_stack_entries = 42; - num_ls_stack_entries = 42; - break; - } - - if ((rdev->family == CHIP_CEDAR) || - (rdev->family == CHIP_PALM) || - (rdev->family == CHIP_CAICOS)) - sq_config = 0; - else - sq_config = VC_ENABLE; - - sq_config |= (EXPORT_SRC_C | - CS_PRIO(0) | - LS_PRIO(0) | - HS_PRIO(0) | - PS_PRIO(0) | - VS_PRIO(1) | - GS_PRIO(2) | - ES_PRIO(3)); - - sq_gpr_resource_mgmt_1 = (NUM_PS_GPRS(num_ps_gprs) | - NUM_VS_GPRS(num_vs_gprs) | - NUM_CLAUSE_TEMP_GPRS(num_temp_gprs)); - sq_gpr_resource_mgmt_2 = (NUM_GS_GPRS(num_gs_gprs) | - NUM_ES_GPRS(num_es_gprs)); - sq_gpr_resource_mgmt_3 = (NUM_HS_GPRS(num_hs_gprs) | - NUM_LS_GPRS(num_ls_gprs)); - sq_thread_resource_mgmt = (NUM_PS_THREADS(num_ps_threads) | - NUM_VS_THREADS(num_vs_threads) | - NUM_GS_THREADS(num_gs_threads) | - NUM_ES_THREADS(num_es_threads)); - sq_thread_resource_mgmt_2 = (NUM_HS_THREADS(num_hs_threads) | - NUM_LS_THREADS(num_ls_threads)); - sq_stack_resource_mgmt_1 = (NUM_PS_STACK_ENTRIES(num_ps_stack_entries) | - NUM_VS_STACK_ENTRIES(num_vs_stack_entries)); - sq_stack_resource_mgmt_2 = (NUM_GS_STACK_ENTRIES(num_gs_stack_entries) | - NUM_ES_STACK_ENTRIES(num_es_stack_entries)); - sq_stack_resource_mgmt_3 = (NUM_HS_STACK_ENTRIES(num_hs_stack_entries) | - NUM_LS_STACK_ENTRIES(num_ls_stack_entries)); - /* set clear context state */ radeon_ring_write(rdev, PACKET3(PACKET3_CLEAR_STATE, 0)); radeon_ring_write(rdev, 0); - /* disable dyn gprs */ - radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 1)); - radeon_ring_write(rdev, (SQ_DYN_GPR_CNTL_PS_FLUSH_REQ - PACKET3_SET_CONFIG_REG_START) >> 2); - radeon_ring_write(rdev, 0); + if (rdev->family < CHIP_CAYMAN) { + switch (rdev->family) { + case CHIP_CEDAR: + default: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 96; + num_vs_threads = 16; + num_gs_threads = 16; + num_es_threads = 16; + num_hs_threads = 16; + num_ls_threads = 16; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_REDWOOD: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_JUNIPER: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 85; + num_vs_stack_entries = 85; + num_gs_stack_entries = 85; + num_es_stack_entries = 85; + num_hs_stack_entries = 85; + num_ls_stack_entries = 85; + break; + case CHIP_CYPRESS: + case CHIP_HEMLOCK: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 85; + num_vs_stack_entries = 85; + num_gs_stack_entries = 85; + num_es_stack_entries = 85; + num_hs_stack_entries = 85; + num_ls_stack_entries = 85; + break; + case CHIP_PALM: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 96; + num_vs_threads = 16; + num_gs_threads = 16; + num_es_threads = 16; + num_hs_threads = 16; + num_ls_threads = 16; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_BARTS: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 85; + num_vs_stack_entries = 85; + num_gs_stack_entries = 85; + num_es_stack_entries = 85; + num_hs_stack_entries = 85; + num_ls_stack_entries = 85; + break; + case CHIP_TURKS: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_CAICOS: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 10; + num_gs_threads = 10; + num_es_threads = 10; + num_hs_threads = 10; + num_ls_threads = 10; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + } - /* SQ config */ - radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 11)); - radeon_ring_write(rdev, (SQ_CONFIG - PACKET3_SET_CONFIG_REG_START) >> 2); - radeon_ring_write(rdev, sq_config); - radeon_ring_write(rdev, sq_gpr_resource_mgmt_1); - radeon_ring_write(rdev, sq_gpr_resource_mgmt_2); - radeon_ring_write(rdev, sq_gpr_resource_mgmt_3); - radeon_ring_write(rdev, 0); - radeon_ring_write(rdev, 0); - radeon_ring_write(rdev, sq_thread_resource_mgmt); - radeon_ring_write(rdev, sq_thread_resource_mgmt_2); - radeon_ring_write(rdev, sq_stack_resource_mgmt_1); - radeon_ring_write(rdev, sq_stack_resource_mgmt_2); - radeon_ring_write(rdev, sq_stack_resource_mgmt_3); + if ((rdev->family == CHIP_CEDAR) || + (rdev->family == CHIP_PALM) || + (rdev->family == CHIP_CAICOS)) + sq_config = 0; + else + sq_config = VC_ENABLE; + + sq_config |= (EXPORT_SRC_C | + CS_PRIO(0) | + LS_PRIO(0) | + HS_PRIO(0) | + PS_PRIO(0) | + VS_PRIO(1) | + GS_PRIO(2) | + ES_PRIO(3)); + + sq_gpr_resource_mgmt_1 = (NUM_PS_GPRS(num_ps_gprs) | + NUM_VS_GPRS(num_vs_gprs) | + NUM_CLAUSE_TEMP_GPRS(num_temp_gprs)); + sq_gpr_resource_mgmt_2 = (NUM_GS_GPRS(num_gs_gprs) | + NUM_ES_GPRS(num_es_gprs)); + sq_gpr_resource_mgmt_3 = (NUM_HS_GPRS(num_hs_gprs) | + NUM_LS_GPRS(num_ls_gprs)); + sq_thread_resource_mgmt = (NUM_PS_THREADS(num_ps_threads) | + NUM_VS_THREADS(num_vs_threads) | + NUM_GS_THREADS(num_gs_threads) | + NUM_ES_THREADS(num_es_threads)); + sq_thread_resource_mgmt_2 = (NUM_HS_THREADS(num_hs_threads) | + NUM_LS_THREADS(num_ls_threads)); + sq_stack_resource_mgmt_1 = (NUM_PS_STACK_ENTRIES(num_ps_stack_entries) | + NUM_VS_STACK_ENTRIES(num_vs_stack_entries)); + sq_stack_resource_mgmt_2 = (NUM_GS_STACK_ENTRIES(num_gs_stack_entries) | + NUM_ES_STACK_ENTRIES(num_es_stack_entries)); + sq_stack_resource_mgmt_3 = (NUM_HS_STACK_ENTRIES(num_hs_stack_entries) | + NUM_LS_STACK_ENTRIES(num_ls_stack_entries)); + + /* disable dyn gprs */ + radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(rdev, (SQ_DYN_GPR_CNTL_PS_FLUSH_REQ - PACKET3_SET_CONFIG_REG_START) >> 2); + radeon_ring_write(rdev, 0); + + /* SQ config */ + radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 11)); + radeon_ring_write(rdev, (SQ_CONFIG - PACKET3_SET_CONFIG_REG_START) >> 2); + radeon_ring_write(rdev, sq_config); + radeon_ring_write(rdev, sq_gpr_resource_mgmt_1); + radeon_ring_write(rdev, sq_gpr_resource_mgmt_2); + radeon_ring_write(rdev, sq_gpr_resource_mgmt_3); + radeon_ring_write(rdev, 0); + radeon_ring_write(rdev, 0); + radeon_ring_write(rdev, sq_thread_resource_mgmt); + radeon_ring_write(rdev, sq_thread_resource_mgmt_2); + radeon_ring_write(rdev, sq_stack_resource_mgmt_1); + radeon_ring_write(rdev, sq_stack_resource_mgmt_2); + radeon_ring_write(rdev, sq_stack_resource_mgmt_3); + } /* CONTEXT_CONTROL */ radeon_ring_write(rdev, 0xc0012800); @@ -570,7 +573,10 @@ int evergreen_blit_init(struct radeon_device *rdev) mutex_init(&rdev->r600_blit.mutex); rdev->r600_blit.state_offset = 0; - rdev->r600_blit.state_len = evergreen_default_size; + if (rdev->family < CHIP_CAYMAN) + rdev->r600_blit.state_len = evergreen_default_size; + else + rdev->r600_blit.state_len = cayman_default_size; dwords = rdev->r600_blit.state_len; while (dwords & 0xf) { @@ -582,11 +588,17 @@ int evergreen_blit_init(struct radeon_device *rdev) obj_size = ALIGN(obj_size, 256); rdev->r600_blit.vs_offset = obj_size; - obj_size += evergreen_vs_size * 4; + if (rdev->family < CHIP_CAYMAN) + obj_size += evergreen_vs_size * 4; + else + obj_size += cayman_vs_size * 4; obj_size = ALIGN(obj_size, 256); rdev->r600_blit.ps_offset = obj_size; - obj_size += evergreen_ps_size * 4; + if (rdev->family < CHIP_CAYMAN) + obj_size += evergreen_ps_size * 4; + else + obj_size += cayman_ps_size * 4; obj_size = ALIGN(obj_size, 256); r = radeon_bo_create(rdev, obj_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, @@ -609,16 +621,29 @@ int evergreen_blit_init(struct radeon_device *rdev) return r; } - memcpy_toio(ptr + rdev->r600_blit.state_offset, - evergreen_default_state, rdev->r600_blit.state_len * 4); - - if (num_packet2s) - memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), - packet2s, num_packet2s * 4); - for (i = 0; i < evergreen_vs_size; i++) - *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(evergreen_vs[i]); - for (i = 0; i < evergreen_ps_size; i++) - *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(evergreen_ps[i]); + if (rdev->family < CHIP_CAYMAN) { + memcpy_toio(ptr + rdev->r600_blit.state_offset, + evergreen_default_state, rdev->r600_blit.state_len * 4); + + if (num_packet2s) + memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), + packet2s, num_packet2s * 4); + for (i = 0; i < evergreen_vs_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(evergreen_vs[i]); + for (i = 0; i < evergreen_ps_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(evergreen_ps[i]); + } else { + memcpy_toio(ptr + rdev->r600_blit.state_offset, + cayman_default_state, rdev->r600_blit.state_len * 4); + + if (num_packet2s) + memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), + packet2s, num_packet2s * 4); + for (i = 0; i < cayman_vs_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(cayman_vs[i]); + for (i = 0; i < cayman_ps_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(cayman_ps[i]); + } radeon_bo_kunmap(rdev->r600_blit.shader_obj); radeon_bo_unreserve(rdev->r600_blit.shader_obj); diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c index a0cc7a5..c023b0a 100644 --- a/drivers/gpu/drm/radeon/ni.c +++ b/drivers/gpu/drm/radeon/ni.c @@ -1387,14 +1387,12 @@ static int cayman_startup(struct radeon_device *rdev) return r; cayman_gpu_init(rdev); -#if 0 - r = cayman_blit_init(rdev); + r = evergreen_blit_init(rdev); if (r) { - cayman_blit_fini(rdev); + evergreen_blit_fini(rdev); rdev->asic->copy = NULL; dev_warn(rdev->dev, "failed blitter (%d) falling back to memcpy\n", r); } -#endif /* allocate wb buffer */ r = radeon_wb_init(rdev); @@ -1452,7 +1450,7 @@ int cayman_resume(struct radeon_device *rdev) int cayman_suspend(struct radeon_device *rdev) { - /* int r; */ + int r; /* FIXME: we should wait for ring to be empty */ cayman_cp_enable(rdev, false); @@ -1461,14 +1459,13 @@ int cayman_suspend(struct radeon_device *rdev) radeon_wb_disable(rdev); cayman_pcie_gart_disable(rdev); -#if 0 /* unpin shaders bo */ r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); if (likely(r == 0)) { radeon_bo_unpin(rdev->r600_blit.shader_obj); radeon_bo_unreserve(rdev->r600_blit.shader_obj); } -#endif + return 0; } @@ -1580,7 +1577,7 @@ int cayman_init(struct radeon_device *rdev) void cayman_fini(struct radeon_device *rdev) { - /* cayman_blit_fini(rdev); */ + evergreen_blit_fini(rdev); cayman_cp_fini(rdev); r600_irq_fini(rdev); radeon_wb_fini(rdev); diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index d948265..b9b3c2a 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -906,9 +906,9 @@ static struct radeon_asic cayman_asic = { .get_vblank_counter = &evergreen_get_vblank_counter, .fence_ring_emit = &r600_fence_ring_emit, .cs_parse = &evergreen_cs_parse, - .copy_blit = NULL, - .copy_dma = NULL, - .copy = NULL, + .copy_blit = &evergreen_copy_blit, + .copy_dma = &evergreen_copy_blit, + .copy = &evergreen_copy_blit, .get_engine_clock = &radeon_atom_get_engine_clock, .set_engine_clock = &radeon_atom_set_engine_clock, .get_memory_clock = &radeon_atom_get_memory_clock, -- cgit v0.10.2 From 457558eda1545c22163574f6dbb883394705e9dd Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 25 May 2011 17:49:54 -0400 Subject: drm/radeon/kms: add missing case for cayman thermal sensor The rest of the code is already in place. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index 86eda1e..aaa19dc 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -487,6 +487,7 @@ static int radeon_hwmon_init(struct radeon_device *rdev) case THERMAL_TYPE_RV6XX: case THERMAL_TYPE_RV770: case THERMAL_TYPE_EVERGREEN: + case THERMAL_TYPE_NI: case THERMAL_TYPE_SUMO: rdev->pm.int_hwmon_dev = hwmon_device_register(rdev->dev); if (IS_ERR(rdev->pm.int_hwmon_dev)) { -- cgit v0.10.2 From 67b3f823ec78d08aea8835bce2655674237abc1d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 25 May 2011 18:45:37 -0400 Subject: drm/radeon/kms: fix thermal sensor reading on juniper Uses a different method than other evergreen asics. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 34cd5a8..8f446aa 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -88,21 +88,39 @@ u32 evergreen_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base) /* get temperature in millidegrees */ int evergreen_get_temp(struct radeon_device *rdev) { - u32 temp = (RREG32(CG_MULT_THERMAL_STATUS) & ASIC_T_MASK) >> - ASIC_T_SHIFT; - u32 actual_temp = 0; - - if (temp & 0x400) - actual_temp = -256; - else if (temp & 0x200) - actual_temp = 255; - else if (temp & 0x100) { - actual_temp = temp & 0x1ff; - actual_temp |= ~0x1ff; - } else - actual_temp = temp & 0xff; + u32 temp, toffset, actual_temp = 0; + + if (rdev->family == CHIP_JUNIPER) { + toffset = (RREG32(CG_THERMAL_CTRL) & TOFFSET_MASK) >> + TOFFSET_SHIFT; + temp = (RREG32(CG_TS0_STATUS) & TS0_ADC_DOUT_MASK) >> + TS0_ADC_DOUT_SHIFT; + + if (toffset & 0x100) + actual_temp = temp / 2 - (0x200 - toffset); + else + actual_temp = temp / 2 + toffset; + + actual_temp = actual_temp * 1000; + + } else { + temp = (RREG32(CG_MULT_THERMAL_STATUS) & ASIC_T_MASK) >> + ASIC_T_SHIFT; + + if (temp & 0x400) + actual_temp = -256; + else if (temp & 0x200) + actual_temp = 255; + else if (temp & 0x100) { + actual_temp = temp & 0x1ff; + actual_temp |= ~0x1ff; + } else + actual_temp = temp & 0xff; + + actual_temp = (actual_temp * 1000) / 2; + } - return (actual_temp * 1000) / 2; + return actual_temp; } int sumo_get_temp(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h index 75b57e3..103250f 100644 --- a/drivers/gpu/drm/radeon/evergreend.h +++ b/drivers/gpu/drm/radeon/evergreend.h @@ -168,10 +168,16 @@ #define SE_DB_BUSY (1 << 30) #define SE_CB_BUSY (1 << 31) /* evergreen */ +#define CG_THERMAL_CTRL 0x72c +#define TOFFSET_MASK 0x00003FE0 +#define TOFFSET_SHIFT 5 #define CG_MULT_THERMAL_STATUS 0x740 #define ASIC_T(x) ((x) << 16) -#define ASIC_T_MASK 0x7FF0000 +#define ASIC_T_MASK 0x07FF0000 #define ASIC_T_SHIFT 16 +#define CG_TS0_STATUS 0x760 +#define TS0_ADC_DOUT_MASK 0x000003FF +#define TS0_ADC_DOUT_SHIFT 0 /* APU */ #define CG_THERMAL_STATUS 0x678 -- cgit v0.10.2 From 2f2f96d1ea72e6602ae440ac2867a6004edb37a1 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 25 May 2011 12:51:44 -0400 Subject: drm/radeon/kms: clean up the radeon kms Kconfig - no longer in staging - all radeons supported Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/Kconfig b/drivers/gpu/drm/radeon/Kconfig index 9746fee..ea92bbe 100644 --- a/drivers/gpu/drm/radeon/Kconfig +++ b/drivers/gpu/drm/radeon/Kconfig @@ -28,11 +28,4 @@ config DRM_RADEON_KMS The kernel will also perform security check on command stream provided by the user, we want to catch and forbid any illegal use of the GPU such as DMA into random system memory or into memory - not owned by the process supplying the command stream. This part - of the code is still incomplete and this why we propose that patch - as a staging driver addition, future security might forbid current - experimental userspace to run. - - This code support the following hardware : R1XX,R2XX,R3XX,R4XX,R5XX - (radeon up to X1950). Works is underway to provide support for R6XX, - R7XX and newer hardware (radeon from HD2XXX to HD4XXX). + not owned by the process supplying the command stream. -- cgit v0.10.2 From 9e3bb6b6f6a0c535eb053fbf0005a8e79e053374 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 24 May 2011 07:51:27 +0200 Subject: KVM: add missing void __user * cast to access_ok() call fa3d315a "KVM: Validate userspace_addr of memslot when registered" introduced this new warning onn s390: kvm_main.c: In function '__kvm_set_memory_region': kvm_main.c:654:7: warning: passing argument 1 of '__access_ok' makes pointer from integer without a cast arch/s390/include/asm/uaccess.h:53:19: note: expected 'const void *' but argument is of type '__u64' Add the missing cast to get rid of it again... Cc: Takuya Yoshikawa Signed-off-by: Heiko Carstens Signed-off-by: Avi Kivity diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 22cdb96..383f492 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -651,7 +651,9 @@ int __kvm_set_memory_region(struct kvm *kvm, /* We can read the guest memory with __xxx_user() later on. */ if (user_alloc && ((mem->userspace_addr & (PAGE_SIZE - 1)) || - !access_ok(VERIFY_WRITE, mem->userspace_addr, mem->memory_size))) + !access_ok(VERIFY_WRITE, + (void __user *)(unsigned long)mem->userspace_addr, + mem->memory_size))) goto out; if (mem->slot >= KVM_MEMORY_SLOTS + KVM_PRIVATE_MEM_SLOTS) goto out; -- cgit v0.10.2 From 5c699d7d3f94ee1dd934edea889b32f8279a4e65 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Thu, 26 May 2011 11:49:16 +0300 Subject: HID: hiddev: fix use after free in hiddev_release There are a couple use after free bugs here. Signed-off-by: Dan Carpenter [jkosina@suse.cz: removed already fixed hunk] Signed-off-by: Jiri Kosina diff --git a/drivers/hid/usbhid/hiddev.c b/drivers/hid/usbhid/hiddev.c index 4985f485..7c1188b 100644 --- a/drivers/hid/usbhid/hiddev.c +++ b/drivers/hid/usbhid/hiddev.c @@ -248,12 +248,15 @@ static int hiddev_release(struct inode * inode, struct file * file) usbhid_close(list->hiddev->hid); usbhid_put_power(list->hiddev->hid); } else { + mutex_unlock(&list->hiddev->existancelock); kfree(list->hiddev); + kfree(list); + return 0; } } - kfree(list); mutex_unlock(&list->hiddev->existancelock); + kfree(list); return 0; } -- cgit v0.10.2 From 0f7e4c33eb2c40b1e9cc24d2eab6de5921bc619c Mon Sep 17 00:00:00 2001 From: Kohei Kaigai Date: Thu, 26 May 2011 14:59:25 -0400 Subject: selinux: fix case of names with whitespace/multibytes on /selinux/create I submit the patch again, according to patch submission convension. This patch enables to accept percent-encoded object names as forth argument of /selinux/create interface to avoid possible bugs when we give an object name including whitespace or multibutes. E.g) if and when a userspace object manager tries to create a new object named as "resolve.conf but fake", it shall give this name as the forth argument of the /selinux/create. But sscanf() logic in kernel space fetches only the part earlier than the first whitespace. In this case, selinux may unexpectedly answer a default security context configured to "resolve.conf", but it is bug. Although I could not test this patch on named TYPE_TRANSITION rules actually, But debug printk() message seems to me the logic works correctly. I assume the libselinux provides an interface to apply this logic transparently, so nothing shall not be changed from the viewpoint of application. Signed-off-by: KaiGai Kohei Signed-off-by: Eric Paris diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index fde4e9d..1948904 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -29,6 +29,7 @@ #include #include #include +#include /* selinuxfs pseudo filesystem for exporting the security policy API. Based on the proc code and the fs/nfsd/nfsctl.c code. */ @@ -751,6 +752,14 @@ out: return length; } +static inline int hexcode_to_int(int code) { + if (code == '\0' || !isxdigit(code)) + return -1; + if (isdigit(code)) + return code - '0'; + return tolower(code) - 'a' + 10; +} + static ssize_t sel_write_create(struct file *file, char *buf, size_t size) { char *scon = NULL, *tcon = NULL; @@ -785,8 +794,34 @@ static ssize_t sel_write_create(struct file *file, char *buf, size_t size) nargs = sscanf(buf, "%s %s %hu %s", scon, tcon, &tclass, namebuf); if (nargs < 3 || nargs > 4) goto out; - if (nargs == 4) + if (nargs == 4) { + /* + * If and when the name of new object to be queried contains + * either whitespace or multibyte characters, they shall be + * encoded based on the percentage-encoding rule. + * If not encoded, the sscanf logic picks up only left-half + * of the supplied name; splitted by a whitespace unexpectedly. + */ + char *r, *w; + int c1, c2; + + r = w = namebuf; + do { + c1 = *r++; + if (c1 == '+') + c1 = ' '; + else if (c1 == '%') { + if ((c1 = hexcode_to_int(*r++)) < 0) + goto out; + if ((c2 = hexcode_to_int(*r++)) < 0) + goto out; + c1 = (c1 << 4) | c2; + } + *w++ = c1; + } while (c1 != '\0'); + objname = namebuf; + } length = security_context_to_sid(scon, strlen(scon) + 1, &ssid); if (length) -- cgit v0.10.2 From 4a8fcc2cc34b7546584bf9ebd8f0caf433e4736e Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Thu, 19 May 2011 20:19:09 -0700 Subject: [SCSI] target: Fix multi task->task_sg[] chaining logic bug This patch fixes a bug in transport_do_task_sg_chain() used by HW target mode modules with sg_chain() to provide a single sg_next() walkable memory layout for use with pci_map_sg() and friends. This patch addresses an issue with mapping multiple small block max_sector tasks across multiple struct se_task->task_sg[] mappings for HW target mode operation. This was causing OOPs with (cmd->t_task->t_tasks_no > 1) I/O traffic for HW target drivers using transport_do_task_sg_chain(), and has been tested so far with tcm_fc(openfcoe), tcm_qla2xxx, and ib_srpt fabrics with t_tasks_no > 1 IBLOCK backends using a smaller max_sectors to trigger the original issue. Signed-off-by: Nicholas Bellinger Acked-by: Kiran Patil Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 9583b23..fefe10a 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -4776,18 +4776,20 @@ void transport_do_task_sg_chain(struct se_cmd *cmd) sg_end_cur->page_link &= ~0x02; sg_chain(sg_head, task_sg_num, sg_head_cur); - sg_count += (task->task_sg_num + 1); - } else sg_count += task->task_sg_num; + task_sg_num = (task->task_sg_num + 1); + } else { + sg_chain(sg_head, task_sg_num, sg_head_cur); + sg_count += task->task_sg_num; + task_sg_num = task->task_sg_num; + } sg_head = sg_head_cur; sg_link = sg_link_cur; - task_sg_num = task->task_sg_num; continue; } sg_head = sg_first = &task->task_sg[0]; sg_link = &task->task_sg[task->task_sg_num]; - task_sg_num = task->task_sg_num; /* * Check for single task.. */ @@ -4798,9 +4800,12 @@ void transport_do_task_sg_chain(struct se_cmd *cmd) */ sg_end = &task->task_sg[task->task_sg_num - 1]; sg_end->page_link &= ~0x02; - sg_count += (task->task_sg_num + 1); - } else sg_count += task->task_sg_num; + task_sg_num = (task->task_sg_num + 1); + } else { + sg_count += task->task_sg_num; + task_sg_num = task->task_sg_num; + } } /* * Setup the starting pointer and total t_tasks_sg_linked_no including @@ -4809,21 +4814,20 @@ void transport_do_task_sg_chain(struct se_cmd *cmd) T_TASK(cmd)->t_tasks_sg_chained = sg_first; T_TASK(cmd)->t_tasks_sg_chained_no = sg_count; - DEBUG_CMD_M("Setup T_TASK(cmd)->t_tasks_sg_chained: %p and" - " t_tasks_sg_chained_no: %u\n", T_TASK(cmd)->t_tasks_sg_chained, + DEBUG_CMD_M("Setup cmd: %p T_TASK(cmd)->t_tasks_sg_chained: %p and" + " t_tasks_sg_chained_no: %u\n", cmd, T_TASK(cmd)->t_tasks_sg_chained, T_TASK(cmd)->t_tasks_sg_chained_no); for_each_sg(T_TASK(cmd)->t_tasks_sg_chained, sg, T_TASK(cmd)->t_tasks_sg_chained_no, i) { - DEBUG_CMD_M("SG: %p page: %p length: %d offset: %d\n", - sg, sg_page(sg), sg->length, sg->offset); + DEBUG_CMD_M("SG[%d]: %p page: %p length: %d offset: %d, magic: 0x%08x\n", + i, sg, sg_page(sg), sg->length, sg->offset, sg->sg_magic); if (sg_is_chain(sg)) DEBUG_CMD_M("SG: %p sg_is_chain=1\n", sg); if (sg_is_last(sg)) DEBUG_CMD_M("SG: %p sg_is_last=1\n", sg); } - } EXPORT_SYMBOL(transport_do_task_sg_chain); -- cgit v0.10.2 From 1e7de68c57daf75ec4b721f101f88cccf029e38c Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Thu, 19 May 2011 20:19:10 -0700 Subject: [SCSI] target: Fix interrupt context bug with stats_lock and core_tmr_alloc_req This patch fixes two bugs wrt to the interrupt context usage of target core with HW target mode drivers. It first converts the usage of struct se_device->stats_lock in transport_get_lun_for_cmd() and core_tmr_lun_reset() to properly use spin_lock_irq() to address an BUG with CONFIG_LOCKDEP_SUPPORT=y enabled. This patch also adds a 'in_interrupt()' check to allow GFP_ATOMIC usage from core_tmr_alloc_req() to fix a 'sleeping in interrupt context' BUG with HW target fabrics that require this logic to function. Signed-off-by: Nicholas Bellinger Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index d25e208..fc10ed4 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -150,13 +150,13 @@ out: { struct se_device *dev = se_lun->lun_se_dev; - spin_lock(&dev->stats_lock); + spin_lock_irq(&dev->stats_lock); dev->num_cmds++; if (se_cmd->data_direction == DMA_TO_DEVICE) dev->write_bytes += se_cmd->data_length; else if (se_cmd->data_direction == DMA_FROM_DEVICE) dev->read_bytes += se_cmd->data_length; - spin_unlock(&dev->stats_lock); + spin_unlock_irq(&dev->stats_lock); } /* diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c index 4a10983..59b8b9c 100644 --- a/drivers/target/target_core_tmr.c +++ b/drivers/target/target_core_tmr.c @@ -55,7 +55,8 @@ struct se_tmr_req *core_tmr_alloc_req( { struct se_tmr_req *tmr; - tmr = kmem_cache_zalloc(se_tmr_req_cache, GFP_KERNEL); + tmr = kmem_cache_zalloc(se_tmr_req_cache, (in_interrupt()) ? + GFP_ATOMIC : GFP_KERNEL); if (!(tmr)) { printk(KERN_ERR "Unable to allocate struct se_tmr_req\n"); return ERR_PTR(-ENOMEM); @@ -398,9 +399,9 @@ int core_tmr_lun_reset( printk(KERN_INFO "LUN_RESET: SCSI-2 Released reservation\n"); } - spin_lock(&dev->stats_lock); + spin_lock_irq(&dev->stats_lock); dev->num_resets++; - spin_unlock(&dev->stats_lock); + spin_unlock_irq(&dev->stats_lock); DEBUG_LR("LUN_RESET: %s for [%s] Complete\n", (preempt_and_abort_list) ? "Preempt" : "TMR", -- cgit v0.10.2 From 42c6951e2f7a665bcb57b92fe3f806ba48152c0e Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Thu, 19 May 2011 20:19:11 -0700 Subject: [SCSI] target: Fix bug with task_sg chained transport_free_dev_tasks release This patch addresses a bug in the target core release path for HW operation where transport_free_dev_tasks() was incorrectly being called from transport_lun_remove_cmd() while releasing a se_cmd reference and calling struct target_core_fabric_ops->queue_data_in(). This would result in a OOPs with HW target mode when the release of se_task->task_sg[] would happen before pci_unmap_sg() can be called in HW target mode fabric module code. This patch addresses the issue by moving transport_free_dev_tasks() from transport_lun_remove_cmd() into transport_generic_free_cmd(), and adding TRANSPORT_FREE_CMD_INTR and transport_generic_free_cmd_intr() to allow se_cmd descriptor release to happen fromfrom within transport_processing_thread() process context when release of se_cmd is not possible from HW interrupt context. Signed-off-by: Nicholas Bellinger Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index fefe10a..3eeb3e2 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -762,7 +762,6 @@ static void transport_lun_remove_cmd(struct se_cmd *cmd) transport_all_task_dev_remove_state(cmd); spin_unlock_irqrestore(&T_TASK(cmd)->t_state_lock, flags); - transport_free_dev_tasks(cmd); check_lun: spin_lock_irqsave(&lun->lun_cmd_lock, flags); @@ -2058,6 +2057,13 @@ int transport_generic_handle_tmr( } EXPORT_SYMBOL(transport_generic_handle_tmr); +void transport_generic_free_cmd_intr( + struct se_cmd *cmd) +{ + transport_add_cmd_to_queue(cmd, TRANSPORT_FREE_CMD_INTR); +} +EXPORT_SYMBOL(transport_generic_free_cmd_intr); + static int transport_stop_tasks_for_cmd(struct se_cmd *cmd) { struct se_task *task, *task_tmp; @@ -5301,6 +5307,8 @@ void transport_generic_free_cmd( if (wait_for_tasks && cmd->transport_wait_for_tasks) cmd->transport_wait_for_tasks(cmd, 0, 0); + transport_free_dev_tasks(cmd); + transport_generic_remove(cmd, release_to_pool, session_reinstatement); } @@ -6136,6 +6144,9 @@ get_cmd: case TRANSPORT_REMOVE: transport_generic_remove(cmd, 1, 0); break; + case TRANSPORT_FREE_CMD_INTR: + transport_generic_free_cmd(cmd, 0, 1, 0); + break; case TRANSPORT_PROCESS_TMR: transport_generic_do_tmr(cmd); break; diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index 1d3b5b2..561ac99 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -98,6 +98,7 @@ enum transport_state_table { TRANSPORT_REMOVE = 14, TRANSPORT_FREE = 15, TRANSPORT_NEW_CMD_MAP = 16, + TRANSPORT_FREE_CMD_INTR = 17, }; /* Used for struct se_cmd->se_cmd_flags */ diff --git a/include/target/target_core_transport.h b/include/target/target_core_transport.h index 59aa464..24a1c6c 100644 --- a/include/target/target_core_transport.h +++ b/include/target/target_core_transport.h @@ -172,6 +172,7 @@ extern int transport_generic_handle_cdb_map(struct se_cmd *); extern int transport_generic_handle_data(struct se_cmd *); extern void transport_new_cmd_failure(struct se_cmd *); extern int transport_generic_handle_tmr(struct se_cmd *); +extern void transport_generic_free_cmd_intr(struct se_cmd *); extern void __transport_stop_task_timer(struct se_task *, unsigned long *); extern unsigned char transport_asciihex_to_binaryhex(unsigned char val[2]); extern int transport_generic_map_mem_to_cmd(struct se_cmd *cmd, struct scatterlist *, u32, -- cgit v0.10.2 From ccf4d680f80941f0073a9bc6a5e0ed41496b46e7 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Thu, 19 May 2011 20:19:12 -0700 Subject: [SCSI] target: Fix task->task_execute_queue=1 clear bug + LUN_RESET OOPs This patch fixes a bug where task->task_execute_queue=1 was not being cleared once se_task had been removed from se_device->execute_task_list, resulting in an OOPs in core_tmr_lun_reset() for the task->task_active=0 case where transport_remove_task_from_execute_queue() was incorrectly being called. This patch fixes two cases in transport_get_task_from_execute_queue() and transport_remove_task_from_execute_queue() to properly clear task->task_execute_queue=0 once list_del(&task->t_execute_list) has been called. It also adds an explict check in transport_remove_task_from_execute_queue() to dump_stack + return if called with task->task_execute_queue=0. Signed-off-by: Nicholas Bellinger Cc: stable@kernel.org Signed-off-by: James Bottomley diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 3eeb3e2..beaf8fa 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -1194,6 +1194,7 @@ transport_get_task_from_execute_queue(struct se_device *dev) break; list_del(&task->t_execute_list); + atomic_set(&task->task_execute_queue, 0); atomic_dec(&dev->execute_tasks); return task; @@ -1209,8 +1210,14 @@ void transport_remove_task_from_execute_queue( { unsigned long flags; + if (atomic_read(&task->task_execute_queue) == 0) { + dump_stack(); + return; + } + spin_lock_irqsave(&dev->execute_task_lock, flags); list_del(&task->t_execute_list); + atomic_set(&task->task_execute_queue, 0); atomic_dec(&dev->execute_tasks); spin_unlock_irqrestore(&dev->execute_task_lock, flags); } -- cgit v0.10.2 From 1078da163fd2c1d3cba76cd4018e897a4f98efc5 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Thu, 19 May 2011 20:19:13 -0700 Subject: [SCSI] target: Convert REPORT_LUNs to use int_to_scsilun This patch converts transport_core_report_lun_response() to use drivers/scsi/scsi_scan.c:int_to_scsilun instead of using the struct target_core_fabric_ops->pack_lun() fabric provided API vector. It also removes the tfo->pack_lun check from target_fabric_tf_ops_check() and removes from struct target_core_fabric_ops->pack_lun() from target_core_fabric_ops.h, and the following mainline scsi-misc fabric modules: *) tcm_loop: Drop tcm_loop_pack_lun() usage *) tcm_fc: Drop ft_pack_lun() usage Reported-by: Mike Christie Signed-off-by: Nicholas A. Bellinger Signed-off-by: James Bottomley diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c index aed4e46..09681ba 100644 --- a/drivers/target/loopback/tcm_loop.c +++ b/drivers/target/loopback/tcm_loop.c @@ -939,18 +939,6 @@ static u16 tcm_loop_get_fabric_sense_len(void) return 0; } -static u64 tcm_loop_pack_lun(unsigned int lun) -{ - u64 result; - - /* LSB of lun into byte 1 big-endian */ - result = ((lun & 0xff) << 8); - /* use flat space addressing method */ - result |= 0x40 | ((lun >> 8) & 0x3f); - - return cpu_to_le64(result); -} - static char *tcm_loop_dump_proto_id(struct tcm_loop_hba *tl_hba) { switch (tl_hba->tl_proto_id) { @@ -1481,7 +1469,6 @@ static int tcm_loop_register_configfs(void) fabric->tf_ops.set_fabric_sense_len = &tcm_loop_set_fabric_sense_len; fabric->tf_ops.get_fabric_sense_len = &tcm_loop_get_fabric_sense_len; fabric->tf_ops.is_state_remove = &tcm_loop_is_state_remove; - fabric->tf_ops.pack_lun = &tcm_loop_pack_lun; tf_cg = &fabric->tf_group; /* diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c index a5f44a6..ee6fad9 100644 --- a/drivers/target/target_core_configfs.c +++ b/drivers/target/target_core_configfs.c @@ -497,10 +497,6 @@ static int target_fabric_tf_ops_check( printk(KERN_ERR "Missing tfo->is_state_remove()\n"); return -EINVAL; } - if (!(tfo->pack_lun)) { - printk(KERN_ERR "Missing tfo->pack_lun()\n"); - return -EINVAL; - } /* * We at least require tfo->fabric_make_wwn(), tfo->fabric_drop_wwn() * tfo->fabric_make_tpg() and tfo->fabric_drop_tpg() in diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index fc10ed4..8407f9c 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include @@ -658,8 +659,7 @@ int transport_core_report_lun_response(struct se_cmd *se_cmd) struct se_session *se_sess = SE_SESS(se_cmd); struct se_task *se_task; unsigned char *buf = (unsigned char *)T_TASK(se_cmd)->t_task_buf; - u32 cdb_offset = 0, lun_count = 0, offset = 8; - u64 i, lun; + u32 cdb_offset = 0, lun_count = 0, offset = 8, i; list_for_each_entry(se_task, &T_TASK(se_cmd)->t_task_list, t_list) break; @@ -675,15 +675,7 @@ int transport_core_report_lun_response(struct se_cmd *se_cmd) * a $FABRIC_MOD. In that case, report LUN=0 only. */ if (!(se_sess)) { - lun = 0; - buf[offset++] = ((lun >> 56) & 0xff); - buf[offset++] = ((lun >> 48) & 0xff); - buf[offset++] = ((lun >> 40) & 0xff); - buf[offset++] = ((lun >> 32) & 0xff); - buf[offset++] = ((lun >> 24) & 0xff); - buf[offset++] = ((lun >> 16) & 0xff); - buf[offset++] = ((lun >> 8) & 0xff); - buf[offset++] = (lun & 0xff); + int_to_scsilun(0, (struct scsi_lun *)&buf[offset]); lun_count = 1; goto done; } @@ -703,15 +695,8 @@ int transport_core_report_lun_response(struct se_cmd *se_cmd) if ((cdb_offset + 8) >= se_cmd->data_length) continue; - lun = cpu_to_be64(CMD_TFO(se_cmd)->pack_lun(deve->mapped_lun)); - buf[offset++] = ((lun >> 56) & 0xff); - buf[offset++] = ((lun >> 48) & 0xff); - buf[offset++] = ((lun >> 40) & 0xff); - buf[offset++] = ((lun >> 32) & 0xff); - buf[offset++] = ((lun >> 24) & 0xff); - buf[offset++] = ((lun >> 16) & 0xff); - buf[offset++] = ((lun >> 8) & 0xff); - buf[offset++] = (lun & 0xff); + int_to_scsilun(deve->mapped_lun, (struct scsi_lun *)&buf[offset]); + offset += 8; cdb_offset += 8; } spin_unlock_irq(&SE_NODE_ACL(se_sess)->device_list_lock); diff --git a/drivers/target/tcm_fc/tfc_conf.c b/drivers/target/tcm_fc/tfc_conf.c index fcdbbff..84e868c 100644 --- a/drivers/target/tcm_fc/tfc_conf.c +++ b/drivers/target/tcm_fc/tfc_conf.c @@ -519,13 +519,6 @@ static u32 ft_tpg_get_inst_index(struct se_portal_group *se_tpg) return tpg->index; } -static u64 ft_pack_lun(unsigned int index) -{ - WARN_ON(index >= 256); - /* Caller wants this byte-swapped */ - return cpu_to_le64((index & 0xff) << 8); -} - static struct target_core_fabric_ops ft_fabric_ops = { .get_fabric_name = ft_get_fabric_name, .get_fabric_proto_ident = fc_get_fabric_proto_ident, @@ -564,7 +557,6 @@ static struct target_core_fabric_ops ft_fabric_ops = { .get_fabric_sense_len = ft_get_fabric_sense_len, .set_fabric_sense_len = ft_set_fabric_sense_len, .is_state_remove = ft_is_state_remove, - .pack_lun = ft_pack_lun, /* * Setup function pointers for generic logic in * target_core_fabric_configfs.c diff --git a/include/target/target_core_fabric_ops.h b/include/target/target_core_fabric_ops.h index dc78f77..747e140 100644 --- a/include/target/target_core_fabric_ops.h +++ b/include/target/target_core_fabric_ops.h @@ -77,7 +77,6 @@ struct target_core_fabric_ops { u16 (*set_fabric_sense_len)(struct se_cmd *, u32); u16 (*get_fabric_sense_len)(void); int (*is_state_remove)(struct se_cmd *); - u64 (*pack_lun)(unsigned int); /* * fabric module calls for target_core_fabric_configfs.c */ -- cgit v0.10.2 From 61db1802bf33bf027cd97ba3f79566b2b2fce5c6 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Thu, 19 May 2011 20:19:14 -0700 Subject: [SCSI] target: Convert TASK_ATTR to scsi_tcq.h definitions This patch converts target core and follwing scsi-misc upstream fabric modules to use include/scsi/scsi_tcq.h includes for SIMPLE, HEAD_OF_QUEUE and ORDERED SCSI tasks instead of scsi/libsas.h with TASK_ATTR* *) tcm_loop: Convert tcm_loop_allocate_core_cmd() + tcm_loop_device_reset() to scsi_tcq.h *) tcm_fc: Convert ft_send_cmd() from FCP_PTA_* to scsi_tcq.h Reported-by: Christoph Hellwig Signed-off-by: Nicholas A. Bellinger Signed-off-by: James Bottomley diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c index 09681ba..dee2a2c 100644 --- a/drivers/target/loopback/tcm_loop.c +++ b/drivers/target/loopback/tcm_loop.c @@ -31,7 +31,7 @@ #include #include #include -#include /* For TASK_ATTR_* */ +#include #include #include @@ -95,17 +95,17 @@ static struct se_cmd *tcm_loop_allocate_core_cmd( if (sc->device->tagged_supported) { switch (sc->tag) { case HEAD_OF_QUEUE_TAG: - sam_task_attr = TASK_ATTR_HOQ; + sam_task_attr = MSG_HEAD_TAG; break; case ORDERED_QUEUE_TAG: - sam_task_attr = TASK_ATTR_ORDERED; + sam_task_attr = MSG_ORDERED_TAG; break; default: - sam_task_attr = TASK_ATTR_SIMPLE; + sam_task_attr = MSG_SIMPLE_TAG; break; } } else - sam_task_attr = TASK_ATTR_SIMPLE; + sam_task_attr = MSG_SIMPLE_TAG; /* * Initialize struct se_cmd descriptor from target_core_mod infrastructure @@ -379,7 +379,7 @@ static int tcm_loop_device_reset(struct scsi_cmnd *sc) * Initialize struct se_cmd descriptor from target_core_mod infrastructure */ transport_init_se_cmd(se_cmd, se_tpg->se_tpg_tfo, se_sess, 0, - DMA_NONE, TASK_ATTR_SIMPLE, + DMA_NONE, MSG_SIMPLE_TAG, &tl_cmd->tl_sense_buf[0]); /* * Allocate the LUN_RESET TMR diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c index 7ff6a35..331d423 100644 --- a/drivers/target/target_core_pscsi.c +++ b/drivers/target/target_core_pscsi.c @@ -41,7 +41,7 @@ #include #include #include -#include /* For TASK_ATTR_* */ +#include #include #include @@ -911,7 +911,7 @@ static int pscsi_do_task(struct se_task *task) * descriptor */ blk_execute_rq_nowait(pdv->pdv_sd->request_queue, NULL, pt->pscsi_req, - (task->task_se_cmd->sam_task_attr == TASK_ATTR_HOQ), + (task->task_se_cmd->sam_task_attr == MSG_HEAD_TAG), pscsi_req_done); return PYX_TRANSPORT_SENT_TO_TRANSPORT; diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index beaf8fa..5b1ce81 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -42,7 +42,7 @@ #include #include #include -#include /* For TASK_ATTR_* */ +#include #include #include @@ -1074,7 +1074,7 @@ static inline int transport_add_task_check_sam_attr( * head of the struct se_device->execute_task_list, and task_prev * after that for each subsequent task */ - if (task->task_se_cmd->sam_task_attr == TASK_ATTR_HOQ) { + if (task->task_se_cmd->sam_task_attr == MSG_HEAD_TAG) { list_add(&task->t_execute_list, (task_prev != NULL) ? &task_prev->t_execute_list : @@ -1873,7 +1873,7 @@ static int transport_check_alloc_task_attr(struct se_cmd *cmd) if (SE_DEV(cmd)->dev_task_attr_type != SAM_TASK_ATTR_EMULATED) return 0; - if (cmd->sam_task_attr == TASK_ATTR_ACA) { + if (cmd->sam_task_attr == MSG_ACA_TAG) { DEBUG_STA("SAM Task Attribute ACA" " emulation is not supported\n"); return -1; @@ -2517,7 +2517,7 @@ static inline int transport_execute_task_attr(struct se_cmd *cmd) * Check for the existence of HEAD_OF_QUEUE, and if true return 1 * to allow the passed struct se_cmd list of tasks to the front of the list. */ - if (cmd->sam_task_attr == TASK_ATTR_HOQ) { + if (cmd->sam_task_attr == MSG_HEAD_TAG) { atomic_inc(&SE_DEV(cmd)->dev_hoq_count); smp_mb__after_atomic_inc(); DEBUG_STA("Added HEAD_OF_QUEUE for CDB:" @@ -2525,7 +2525,7 @@ static inline int transport_execute_task_attr(struct se_cmd *cmd) T_TASK(cmd)->t_task_cdb[0], cmd->se_ordered_id); return 1; - } else if (cmd->sam_task_attr == TASK_ATTR_ORDERED) { + } else if (cmd->sam_task_attr == MSG_ORDERED_TAG) { spin_lock(&SE_DEV(cmd)->ordered_cmd_lock); list_add_tail(&cmd->se_ordered_list, &SE_DEV(cmd)->ordered_cmd_list); @@ -3424,7 +3424,7 @@ static int transport_generic_cmd_sequencer( * See spc4r17 section 5.3 */ if (SE_DEV(cmd)->dev_task_attr_type == SAM_TASK_ATTR_EMULATED) - cmd->sam_task_attr = TASK_ATTR_HOQ; + cmd->sam_task_attr = MSG_HEAD_TAG; cmd->se_cmd_flags |= SCF_SCSI_CONTROL_NONSG_IO_CDB; break; case READ_BUFFER: @@ -3632,7 +3632,7 @@ static int transport_generic_cmd_sequencer( * See spc4r17 section 5.3 */ if (SE_DEV(cmd)->dev_task_attr_type == SAM_TASK_ATTR_EMULATED) - cmd->sam_task_attr = TASK_ATTR_HOQ; + cmd->sam_task_attr = MSG_HEAD_TAG; cmd->se_cmd_flags |= SCF_SCSI_CONTROL_NONSG_IO_CDB; break; default: @@ -3790,21 +3790,21 @@ static void transport_complete_task_attr(struct se_cmd *cmd) struct se_cmd *cmd_p, *cmd_tmp; int new_active_tasks = 0; - if (cmd->sam_task_attr == TASK_ATTR_SIMPLE) { + if (cmd->sam_task_attr == MSG_SIMPLE_TAG) { atomic_dec(&dev->simple_cmds); smp_mb__after_atomic_dec(); dev->dev_cur_ordered_id++; DEBUG_STA("Incremented dev->dev_cur_ordered_id: %u for" " SIMPLE: %u\n", dev->dev_cur_ordered_id, cmd->se_ordered_id); - } else if (cmd->sam_task_attr == TASK_ATTR_HOQ) { + } else if (cmd->sam_task_attr == MSG_HEAD_TAG) { atomic_dec(&dev->dev_hoq_count); smp_mb__after_atomic_dec(); dev->dev_cur_ordered_id++; DEBUG_STA("Incremented dev_cur_ordered_id: %u for" " HEAD_OF_QUEUE: %u\n", dev->dev_cur_ordered_id, cmd->se_ordered_id); - } else if (cmd->sam_task_attr == TASK_ATTR_ORDERED) { + } else if (cmd->sam_task_attr == MSG_ORDERED_TAG) { spin_lock(&dev->ordered_cmd_lock); list_del(&cmd->se_ordered_list); atomic_dec(&dev->dev_ordered_sync); @@ -3837,7 +3837,7 @@ static void transport_complete_task_attr(struct se_cmd *cmd) new_active_tasks++; spin_lock(&dev->delayed_cmd_lock); - if (cmd_p->sam_task_attr == TASK_ATTR_ORDERED) + if (cmd_p->sam_task_attr == MSG_ORDERED_TAG) break; } spin_unlock(&dev->delayed_cmd_lock); diff --git a/drivers/target/tcm_fc/tfc_cmd.c b/drivers/target/tcm_fc/tfc_cmd.c index 49e5177..c056a11 100644 --- a/drivers/target/tcm_fc/tfc_cmd.c +++ b/drivers/target/tcm_fc/tfc_cmd.c @@ -35,6 +35,7 @@ #include #include #include +#include #include #include @@ -592,8 +593,25 @@ static void ft_send_cmd(struct ft_cmd *cmd) case FCP_CFL_WRDATA | FCP_CFL_RDDATA: goto err; /* TBD not supported by tcm_fc yet */ } + /* + * Locate the SAM Task Attr from fc_pri_ta + */ + switch (fcp->fc_pri_ta & FCP_PTA_MASK) { + case FCP_PTA_HEADQ: + task_attr = MSG_HEAD_TAG; + break; + case FCP_PTA_ORDERED: + task_attr = MSG_ORDERED_TAG; + break; + case FCP_PTA_ACA: + task_attr = MSG_ACA_TAG; + break; + case FCP_PTA_SIMPLE: /* Fallthrough */ + default: + task_attr = MSG_SIMPLE_TAG; + } + - /* FCP_PTA_ maps 1:1 to TASK_ATTR_ */ task_attr = fcp->fc_pri_ta & FCP_PTA_MASK; data_len = ntohl(fcp->fc_dl); cmd->cdb = fcp->fc_cdb; diff --git a/include/scsi/scsi_tcq.h b/include/scsi/scsi_tcq.h index d6e7994..81dd12e 100644 --- a/include/scsi/scsi_tcq.h +++ b/include/scsi/scsi_tcq.h @@ -9,6 +9,7 @@ #define MSG_SIMPLE_TAG 0x20 #define MSG_HEAD_TAG 0x21 #define MSG_ORDERED_TAG 0x22 +#define MSG_ACA_TAG 0x24 /* unsupported */ #define SCSI_NO_TAG (-1) /* identify no tag in use */ -- cgit v0.10.2 From 27721a52d6c8e33327ec3cae9f730204be99d251 Mon Sep 17 00:00:00 2001 From: Kyungmin Park Date: Mon, 30 May 2011 10:24:47 +0900 Subject: gpio: Fix gpio-exynos4 build fails in mainline After the GPIO driver move, some symbols became selectable when they shouldn't be. Tighten the dependencies. Reported-by: Randy Dunlap Signed-off-by: Grant Likely diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig index 4a7f631..3171997 100644 --- a/drivers/gpio/Kconfig +++ b/drivers/gpio/Kconfig @@ -88,29 +88,29 @@ config GPIO_IT8761E config GPIO_EXYNOS4 bool "Samsung Exynos4 GPIO library support" - default y if CPU_EXYNOS4210 - depends on ARM + default y + depends on ARM && CPU_EXYNOS4210 help Say yes here to support Samsung Exynos4 series SoCs GPIO library config GPIO_PLAT_SAMSUNG bool "Samsung SoCs GPIO library support" - default y if SAMSUNG_GPIOLIB_4BIT - depends on ARM + default y + depends on ARM && SAMSUNG_GPIOLIB_4BIT help Say yes here to support Samsung SoCs GPIO library config GPIO_S5PC100 bool "Samsung S5PC100 GPIO library support" - default y if CPU_S5PC100 - depends on ARM + default y + depends on ARM && CPU_S5PC100 help Say yes here to support Samsung S5PC100 SoCs GPIO library config GPIO_S5PV210 bool "Samsung S5PV210/S5PC110 GPIO library support" - default y if CPU_S5PV210 - depends on ARM + default y + depends on ARM && CPU_S5PV210 help Say yes here to support Samsung S5PV210/S5PC110 SoCs GPIO library -- cgit v0.10.2 From cbf74cea070fa1f705de4712e25d9e56ae6543c7 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Mon, 30 May 2011 16:31:11 +0200 Subject: oprofile, x86: Add comments to IBS LVT offset initialization Adding a comment in the code as IBS LVT setup is not obvious at all ... Signed-off-by: Robert Richter diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index fabf01e..a0bf78a 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -390,7 +390,8 @@ static unsigned int reserve_eilvt_offset(int offset, unsigned int new) /* * If mask=1, the LVT entry does not generate interrupts while mask=0 - * enables the vector. See also the BKDGs. + * enables the vector. See also the BKDGs. Must be called with + * preemption disabled. */ int setup_APIC_eilvt(u8 offset, u8 vector, u8 msg_type, u8 mask) diff --git a/arch/x86/oprofile/op_model_amd.c b/arch/x86/oprofile/op_model_amd.c index 9fd8a56..9cbb710 100644 --- a/arch/x86/oprofile/op_model_amd.c +++ b/arch/x86/oprofile/op_model_amd.c @@ -609,16 +609,21 @@ static int setup_ibs_ctl(int ibs_eilvt_off) return 0; } +/* + * This runs only on the current cpu. We try to find an LVT offset and + * setup the local APIC. For this we must disable preemption. On + * success we initialize all nodes with this offset. This updates then + * the offset in the IBS_CTL per-node msr. The per-core APIC setup of + * the IBS interrupt vector is called from op_amd_setup_ctrs()/op_- + * amd_cpu_shutdown() using the new offset. + */ static int force_ibs_eilvt_setup(void) { int offset; int ret; - /* - * find the next free available EILVT entry, skip offset 0, - * pin search to this cpu - */ preempt_disable(); + /* find the next free available EILVT entry, skip offset 0 */ for (offset = 1; offset < APIC_EILVT_NR_MAX; offset++) { if (get_eilvt(offset)) break; -- cgit v0.10.2 From 1adffbae22332bb558c2a29de19d9aca391869f6 Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Tue, 31 May 2011 19:38:07 +0900 Subject: fat: Fix corrupt inode flags when remove ATTR_SYS flag We are clearly missing '~' in fat_ioctl_set_attributes(). Cc: Reported-by: Dmitry Dmitriev Signed-off-by: OGAWA Hirofumi diff --git a/fs/fat/file.c b/fs/fat/file.c index 7257752..7018e1d 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -102,7 +102,7 @@ static int fat_ioctl_set_attributes(struct file *file, u32 __user *user_attr) if (attr & ATTR_SYS) inode->i_flags |= S_IMMUTABLE; else - inode->i_flags &= S_IMMUTABLE; + inode->i_flags &= ~S_IMMUTABLE; } fat_save_attrs(inode, attr); -- cgit v0.10.2 From f01114cb59d670e9b4f2c335930dd57db96e9360 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 31 May 2011 12:26:55 +0200 Subject: sched: Fix cross-cpu clock sync on remote wakeups Markus reported that commit 317f394160e ("sched: Move the second half of ttwu() to the remote cpu") caused some accounting funnies on his AMD Phenom II X4, such as weird 'top' results. It turns out that this is due to non-synced TSC and the queued remote wakeups stopped coupeling the two relevant cpu clocks, which leads to wakeups seeing time jumps, which in turn lead to skewed runtime stats. Add an explicit call to sched_clock_cpu() to couple the per-cpu clocks to restore the normal flow of time. Reported-and-tested-by: Markus Trippelsdorf Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/1306835745.2353.3.camel@twins Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index cbb3a0e..49cc70b 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2600,6 +2600,7 @@ static void ttwu_queue(struct task_struct *p, int cpu) #if defined(CONFIG_SMP) if (sched_feat(TTWU_QUEUE) && cpu != smp_processor_id()) { + sched_clock_cpu(cpu); /* sync clocks x-cpu */ ttwu_queue_remote(p, cpu); return; } -- cgit v0.10.2 From f339b9dc1f03591761d5d930800db24bc0eda1e1 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 31 May 2011 10:49:20 +0200 Subject: sched: Fix schedstat.nr_wakeups_migrate While looking over the code I found that with the ttwu rework the nr_wakeups_migrate test broke since we now switch cpus prior to calling ttwu_stat(), hence the test is always true. Cure this by passing the migration state in wake_flags. Also move the whole test under CONFIG_SMP, its hard to migrate tasks on UP :-) Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/n/tip-pwwxl7gdqs5676f1d4cx6pj7@git.kernel.org Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 8da84b7..483c1ed 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1063,6 +1063,7 @@ struct sched_domain; */ #define WF_SYNC 0x01 /* waker goes to sleep after wakup */ #define WF_FORK 0x02 /* child wakeup after fork */ +#define WF_MIGRATED 0x04 /* internal use, task got migrated */ #define ENQUEUE_WAKEUP 1 #define ENQUEUE_HEAD 2 diff --git a/kernel/sched.c b/kernel/sched.c index 49cc70b..2fe98ed 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2447,6 +2447,10 @@ ttwu_stat(struct task_struct *p, int cpu, int wake_flags) } rcu_read_unlock(); } + + if (wake_flags & WF_MIGRATED) + schedstat_inc(p, se.statistics.nr_wakeups_migrate); + #endif /* CONFIG_SMP */ schedstat_inc(rq, ttwu_count); @@ -2455,9 +2459,6 @@ ttwu_stat(struct task_struct *p, int cpu, int wake_flags) if (wake_flags & WF_SYNC) schedstat_inc(p, se.statistics.nr_wakeups_sync); - if (cpu != task_cpu(p)) - schedstat_inc(p, se.statistics.nr_wakeups_migrate); - #endif /* CONFIG_SCHEDSTATS */ } @@ -2675,8 +2676,10 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) p->sched_class->task_waking(p); cpu = select_task_rq(p, SD_BALANCE_WAKE, wake_flags); - if (task_cpu(p) != cpu) + if (task_cpu(p) != cpu) { + wake_flags |= WF_MIGRATED; set_task_cpu(p, cpu); + } #endif /* CONFIG_SMP */ ttwu_queue(p, cpu); -- cgit v0.10.2 From 74c355fbdfedd3820046dba4f537876cea54c207 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 30 May 2011 16:48:06 +0200 Subject: perf, cgroups: Fix up for new API Ben changed the cgroup API in commit f780bdb7c1c (cgroups: add per-thread subsystem callbacks) in an incompatible way, but forgot to convert the perf cgroup bits. Avoid compile warnings and runtime splats and convert perf too ;-) Acked-by: Ben Blum Cc: Stephane Eranian Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/1306767651.1200.2990.camel@twins Signed-off-by: Ingo Molnar diff --git a/kernel/events/core.c b/kernel/events/core.c index c09767f..8a15944 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -7394,26 +7394,12 @@ static int __perf_cgroup_move(void *info) return 0; } -static void perf_cgroup_move(struct task_struct *task) +static void +perf_cgroup_attach_task(struct cgroup *cgrp, struct task_struct *task) { task_function_call(task, __perf_cgroup_move, task); } -static void perf_cgroup_attach(struct cgroup_subsys *ss, struct cgroup *cgrp, - struct cgroup *old_cgrp, struct task_struct *task, - bool threadgroup) -{ - perf_cgroup_move(task); - if (threadgroup) { - struct task_struct *c; - rcu_read_lock(); - list_for_each_entry_rcu(c, &task->thread_group, thread_group) { - perf_cgroup_move(c); - } - rcu_read_unlock(); - } -} - static void perf_cgroup_exit(struct cgroup_subsys *ss, struct cgroup *cgrp, struct cgroup *old_cgrp, struct task_struct *task) { @@ -7425,7 +7411,7 @@ static void perf_cgroup_exit(struct cgroup_subsys *ss, struct cgroup *cgrp, if (!(task->flags & PF_EXITING)) return; - perf_cgroup_move(task); + perf_cgroup_attach_task(cgrp, task); } struct cgroup_subsys perf_subsys = { @@ -7434,6 +7420,6 @@ struct cgroup_subsys perf_subsys = { .create = perf_cgroup_create, .destroy = perf_cgroup_destroy, .exit = perf_cgroup_exit, - .attach = perf_cgroup_attach, + .attach_task = perf_cgroup_attach_task, }; #endif /* CONFIG_CGROUP_PERF */ -- cgit v0.10.2 From 046d886db18260d4435c942985392b7ef872337f Mon Sep 17 00:00:00 2001 From: Janusz Krzysztofik Date: Tue, 31 May 2011 00:44:31 +0200 Subject: OMAP: fix compilation error Forward-declare platform_device structure in arch/arm/plat-omap/include/plat/flash.h, otherwise compilation may break with: In file included from arch/arm/mach-omap1/flash.c:15: arch/arm/plat-omap/include/plat/flash.h:14: warning: 'struct platform_device' declared inside parameter list arch/arm/plat-omap/include/plat/flash.h:14: warning: its scope is only this definition or declaration, which is probably not what you want arch/arm/mach-omap1/flash.c:16: warning: 'struct platform_device' declared inside parameter list arch/arm/mach-omap1/flash.c:17: error: conflicting types for 'omap1_set_vpp' arch/arm/plat-omap/include/plat/flash.h:14: error: previous declaration of 'omap1_set_vpp' was here Detected and corrected while building for Amstrad Delta, confirmed with omap1_defconfig. Signed-off-by: Janusz Krzysztofik Signed-off-by: Tony Lindgren diff --git a/arch/arm/plat-omap/include/plat/flash.h b/arch/arm/plat-omap/include/plat/flash.h index 3083195..0d88499 100644 --- a/arch/arm/plat-omap/include/plat/flash.h +++ b/arch/arm/plat-omap/include/plat/flash.h @@ -11,6 +11,7 @@ #include +struct platform_device; extern void omap1_set_vpp(struct platform_device *pdev, int enable); #endif -- cgit v0.10.2 From f9fa1bb9d7a3c9537d6abd4ad47c58fe27862625 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Tue, 31 May 2011 20:55:44 +0800 Subject: ARM: OMAP2: Add missing include of linux/gpio.h I got some build error like below while executing "make omap2plus_defconfig". CC arch/arm/mach-omap2/board-2430sdp.o arch/arm/mach-omap2/board-2430sdp.c: In function 'omap_2430sdp_init': arch/arm/mach-omap2/board-2430sdp.c:247: error: 'GPIOF_OUT_INIT_LOW' undeclared (first use in this function) arch/arm/mach-omap2/board-2430sdp.c:247: error: (Each undeclared identifier is reported only once arch/arm/mach-omap2/board-2430sdp.c:247: error: for each function it appears in.) This patch fixes the build error by include linux/gpio.h instead of mach/gpio.h. Signed-off-by: Axel Lin Cc: Syed Mohammed Khasim Cc: Grazvydas Ignotas Cc: Steve Sakoman Signed-off-by: Tony Lindgren diff --git a/arch/arm/mach-omap2/board-2430sdp.c b/arch/arm/mach-omap2/board-2430sdp.c index d54969b..5de6eac 100644 --- a/arch/arm/mach-omap2/board-2430sdp.c +++ b/arch/arm/mach-omap2/board-2430sdp.c @@ -26,13 +26,13 @@ #include #include #include +#include #include #include #include #include -#include #include #include #include diff --git a/arch/arm/mach-omap2/board-apollon.c b/arch/arm/mach-omap2/board-apollon.c index f3beb8e..b124bdf 100644 --- a/arch/arm/mach-omap2/board-apollon.c +++ b/arch/arm/mach-omap2/board-apollon.c @@ -27,13 +27,13 @@ #include #include #include +#include #include #include #include #include -#include #include #include #include diff --git a/arch/arm/mach-omap2/board-omap3pandora.c b/arch/arm/mach-omap2/board-omap3pandora.c index 1d10736..5ee034a 100644 --- a/arch/arm/mach-omap2/board-omap3pandora.c +++ b/arch/arm/mach-omap2/board-omap3pandora.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include @@ -41,7 +42,6 @@ #include #include -#include #include #include #include diff --git a/arch/arm/mach-omap2/board-overo.c b/arch/arm/mach-omap2/board-overo.c index 1555918..9952c20 100644 --- a/arch/arm/mach-omap2/board-overo.c +++ b/arch/arm/mach-omap2/board-overo.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include #include @@ -45,7 +46,6 @@ #include #include