diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2015-02-10 21:56:56 (GMT) |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2015-02-10 21:56:56 (GMT) |
commit | bdccc4edeb03ad68c55053b0260bdaaac547bbd9 (patch) | |
tree | c9fdce7c6bb13632f5a64925a1fb3fa306b7cfe1 /drivers | |
parent | 98368ab436538103a557fc1f15f54afd8aab6712 (diff) | |
parent | 72978b2fe2f2cdf9f319c6c6dcdbe92b38de2be2 (diff) | |
download | linux-bdccc4edeb03ad68c55053b0260bdaaac547bbd9.tar.xz |
Merge tag 'stable/for-linus-3.20-rc0-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen features and fixes from David Vrabel:
- Reworked handling for foreign (grant mapped) pages to simplify the
code, enable a number of additional use cases and fix a number of
long-standing bugs.
- Prefer the TSC over the Xen PV clock when dom0 (and the TSC is
stable).
- Assorted other cleanup and minor bug fixes.
* tag 'stable/for-linus-3.20-rc0-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip: (25 commits)
xen/manage: Fix USB interaction issues when resuming
xenbus: Add proper handling of XS_ERROR from Xenbus for transactions.
xen/gntdev: provide find_special_page VMA operation
xen/gntdev: mark userspace PTEs as special on x86 PV guests
xen-blkback: safely unmap grants in case they are still in use
xen/gntdev: safely unmap grants in case they are still in use
xen/gntdev: convert priv->lock to a mutex
xen/grant-table: add a mechanism to safely unmap pages that are in use
xen-netback: use foreign page information from the pages themselves
xen: mark grant mapped pages as foreign
xen/grant-table: add helpers for allocating pages
x86/xen: require ballooned pages for grant maps
xen: remove scratch frames for ballooned pages and m2p override
xen/grant-table: pre-populate kernel unmap ops for xen_gnttab_unmap_refs()
mm: add 'foreign' alias for the 'pinned' page flag
mm: provide a find_special_page vma operation
x86/xen: cleanup arch/x86/xen/mmu.c
x86/xen: add some __init annotations in arch/x86/xen/mmu.c
x86/xen: add some __init and static annotations in arch/x86/xen/setup.c
x86/xen: use correct types for addresses in arch/x86/xen/setup.c
...
Diffstat (limited to 'drivers')
-rw-r--r-- | drivers/block/xen-blkback/blkback.c | 177 | ||||
-rw-r--r-- | drivers/block/xen-blkback/common.h | 3 | ||||
-rw-r--r-- | drivers/net/xen-netback/interface.c | 7 | ||||
-rw-r--r-- | drivers/net/xen-netback/netback.c | 106 | ||||
-rw-r--r-- | drivers/xen/balloon.c | 86 | ||||
-rw-r--r-- | drivers/xen/gntdev.c | 143 | ||||
-rw-r--r-- | drivers/xen/grant-table.c | 120 | ||||
-rw-r--r-- | drivers/xen/manage.c | 8 | ||||
-rw-r--r-- | drivers/xen/tmem.c | 2 | ||||
-rw-r--r-- | drivers/xen/xen-scsiback.c | 6 | ||||
-rw-r--r-- | drivers/xen/xenbus/xenbus_dev_frontend.c | 11 |
11 files changed, 385 insertions, 284 deletions
diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c index 63fc7f0..2a04d34 100644 --- a/drivers/block/xen-blkback/blkback.c +++ b/drivers/block/xen-blkback/blkback.c @@ -47,6 +47,7 @@ #include <asm/xen/hypervisor.h> #include <asm/xen/hypercall.h> #include <xen/balloon.h> +#include <xen/grant_table.h> #include "common.h" /* @@ -100,7 +101,7 @@ module_param(log_stats, int, 0644); #define BLKBACK_INVALID_HANDLE (~0) -/* Number of free pages to remove on each call to free_xenballooned_pages */ +/* Number of free pages to remove on each call to gnttab_free_pages */ #define NUM_BATCH_FREE_PAGES 10 static inline int get_free_page(struct xen_blkif *blkif, struct page **page) @@ -111,7 +112,7 @@ static inline int get_free_page(struct xen_blkif *blkif, struct page **page) if (list_empty(&blkif->free_pages)) { BUG_ON(blkif->free_pages_num != 0); spin_unlock_irqrestore(&blkif->free_pages_lock, flags); - return alloc_xenballooned_pages(1, page, false); + return gnttab_alloc_pages(1, page); } BUG_ON(blkif->free_pages_num == 0); page[0] = list_first_entry(&blkif->free_pages, struct page, lru); @@ -151,14 +152,14 @@ static inline void shrink_free_pagepool(struct xen_blkif *blkif, int num) blkif->free_pages_num--; if (++num_pages == NUM_BATCH_FREE_PAGES) { spin_unlock_irqrestore(&blkif->free_pages_lock, flags); - free_xenballooned_pages(num_pages, page); + gnttab_free_pages(num_pages, page); spin_lock_irqsave(&blkif->free_pages_lock, flags); num_pages = 0; } } spin_unlock_irqrestore(&blkif->free_pages_lock, flags); if (num_pages != 0) - free_xenballooned_pages(num_pages, page); + gnttab_free_pages(num_pages, page); } #define vaddr(page) ((unsigned long)pfn_to_kaddr(page_to_pfn(page))) @@ -262,6 +263,17 @@ static void put_persistent_gnt(struct xen_blkif *blkif, atomic_dec(&blkif->persistent_gnt_in_use); } +static void free_persistent_gnts_unmap_callback(int result, + struct gntab_unmap_queue_data *data) +{ + struct completion *c = data->data; + + /* BUG_ON used to reproduce existing behaviour, + but is this the best way to deal with this? */ + BUG_ON(result); + complete(c); +} + static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, unsigned int num) { @@ -269,8 +281,17 @@ static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, struct page *pages[BLKIF_MAX_SEGMENTS_PER_REQUEST]; struct persistent_gnt *persistent_gnt; struct rb_node *n; - int ret = 0; int segs_to_unmap = 0; + struct gntab_unmap_queue_data unmap_data; + struct completion unmap_completion; + + init_completion(&unmap_completion); + + unmap_data.data = &unmap_completion; + unmap_data.done = &free_persistent_gnts_unmap_callback; + unmap_data.pages = pages; + unmap_data.unmap_ops = unmap; + unmap_data.kunmap_ops = NULL; foreach_grant_safe(persistent_gnt, n, root, node) { BUG_ON(persistent_gnt->handle == @@ -285,9 +306,11 @@ static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, if (++segs_to_unmap == BLKIF_MAX_SEGMENTS_PER_REQUEST || !rb_next(&persistent_gnt->node)) { - ret = gnttab_unmap_refs(unmap, NULL, pages, - segs_to_unmap); - BUG_ON(ret); + + unmap_data.count = segs_to_unmap; + gnttab_unmap_refs_async(&unmap_data); + wait_for_completion(&unmap_completion); + put_free_pages(blkif, pages, segs_to_unmap); segs_to_unmap = 0; } @@ -653,18 +676,14 @@ void xen_blkbk_free_caches(struct xen_blkif *blkif) shrink_free_pagepool(blkif, 0 /* All */); } -/* - * Unmap the grant references, and also remove the M2P over-rides - * used in the 'pending_req'. - */ -static void xen_blkbk_unmap(struct xen_blkif *blkif, - struct grant_page *pages[], - int num) +static unsigned int xen_blkbk_unmap_prepare( + struct xen_blkif *blkif, + struct grant_page **pages, + unsigned int num, + struct gnttab_unmap_grant_ref *unmap_ops, + struct page **unmap_pages) { - struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST]; - struct page *unmap_pages[BLKIF_MAX_SEGMENTS_PER_REQUEST]; unsigned int i, invcount = 0; - int ret; for (i = 0; i < num; i++) { if (pages[i]->persistent_gnt != NULL) { @@ -674,21 +693,95 @@ static void xen_blkbk_unmap(struct xen_blkif *blkif, if (pages[i]->handle == BLKBACK_INVALID_HANDLE) continue; unmap_pages[invcount] = pages[i]->page; - gnttab_set_unmap_op(&unmap[invcount], vaddr(pages[i]->page), + gnttab_set_unmap_op(&unmap_ops[invcount], vaddr(pages[i]->page), GNTMAP_host_map, pages[i]->handle); pages[i]->handle = BLKBACK_INVALID_HANDLE; - if (++invcount == BLKIF_MAX_SEGMENTS_PER_REQUEST) { - ret = gnttab_unmap_refs(unmap, NULL, unmap_pages, - invcount); + invcount++; + } + + return invcount; +} + +static void xen_blkbk_unmap_and_respond_callback(int result, struct gntab_unmap_queue_data *data) +{ + struct pending_req* pending_req = (struct pending_req*) (data->data); + struct xen_blkif *blkif = pending_req->blkif; + + /* BUG_ON used to reproduce existing behaviour, + but is this the best way to deal with this? */ + BUG_ON(result); + + put_free_pages(blkif, data->pages, data->count); + make_response(blkif, pending_req->id, + pending_req->operation, pending_req->status); + free_req(blkif, pending_req); + /* + * Make sure the request is freed before releasing blkif, + * or there could be a race between free_req and the + * cleanup done in xen_blkif_free during shutdown. + * + * NB: The fact that we might try to wake up pending_free_wq + * before drain_complete (in case there's a drain going on) + * it's not a problem with our current implementation + * because we can assure there's no thread waiting on + * pending_free_wq if there's a drain going on, but it has + * to be taken into account if the current model is changed. + */ + if (atomic_dec_and_test(&blkif->inflight) && atomic_read(&blkif->drain)) { + complete(&blkif->drain_complete); + } + xen_blkif_put(blkif); +} + +static void xen_blkbk_unmap_and_respond(struct pending_req *req) +{ + struct gntab_unmap_queue_data* work = &req->gnttab_unmap_data; + struct xen_blkif *blkif = req->blkif; + struct grant_page **pages = req->segments; + unsigned int invcount; + + invcount = xen_blkbk_unmap_prepare(blkif, pages, req->nr_pages, + req->unmap, req->unmap_pages); + + work->data = req; + work->done = xen_blkbk_unmap_and_respond_callback; + work->unmap_ops = req->unmap; + work->kunmap_ops = NULL; + work->pages = req->unmap_pages; + work->count = invcount; + + gnttab_unmap_refs_async(&req->gnttab_unmap_data); +} + + +/* + * Unmap the grant references. + * + * This could accumulate ops up to the batch size to reduce the number + * of hypercalls, but since this is only used in error paths there's + * no real need. + */ +static void xen_blkbk_unmap(struct xen_blkif *blkif, + struct grant_page *pages[], + int num) +{ + struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST]; + struct page *unmap_pages[BLKIF_MAX_SEGMENTS_PER_REQUEST]; + unsigned int invcount = 0; + int ret; + + while (num) { + unsigned int batch = min(num, BLKIF_MAX_SEGMENTS_PER_REQUEST); + + invcount = xen_blkbk_unmap_prepare(blkif, pages, batch, + unmap, unmap_pages); + if (invcount) { + ret = gnttab_unmap_refs(unmap, NULL, unmap_pages, invcount); BUG_ON(ret); put_free_pages(blkif, unmap_pages, invcount); - invcount = 0; } - } - if (invcount) { - ret = gnttab_unmap_refs(unmap, NULL, unmap_pages, invcount); - BUG_ON(ret); - put_free_pages(blkif, unmap_pages, invcount); + pages += batch; + num -= batch; } } @@ -982,32 +1075,8 @@ static void __end_block_io_op(struct pending_req *pending_req, int error) * the grant references associated with 'request' and provide * the proper response on the ring. */ - if (atomic_dec_and_test(&pending_req->pendcnt)) { - struct xen_blkif *blkif = pending_req->blkif; - - xen_blkbk_unmap(blkif, - pending_req->segments, - pending_req->nr_pages); - make_response(blkif, pending_req->id, - pending_req->operation, pending_req->status); - free_req(blkif, pending_req); - /* - * Make sure the request is freed before releasing blkif, - * or there could be a race between free_req and the - * cleanup done in xen_blkif_free during shutdown. - * - * NB: The fact that we might try to wake up pending_free_wq - * before drain_complete (in case there's a drain going on) - * it's not a problem with our current implementation - * because we can assure there's no thread waiting on - * pending_free_wq if there's a drain going on, but it has - * to be taken into account if the current model is changed. - */ - if (atomic_dec_and_test(&blkif->inflight) && atomic_read(&blkif->drain)) { - complete(&blkif->drain_complete); - } - xen_blkif_put(blkif); - } + if (atomic_dec_and_test(&pending_req->pendcnt)) + xen_blkbk_unmap_and_respond(pending_req); } /* diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h index f65b807..cc90a84 100644 --- a/drivers/block/xen-blkback/common.h +++ b/drivers/block/xen-blkback/common.h @@ -350,6 +350,9 @@ struct pending_req { struct grant_page *indirect_pages[MAX_INDIRECT_PAGES]; struct seg_buf seg[MAX_INDIRECT_SEGMENTS]; struct bio *biolist[MAX_INDIRECT_SEGMENTS]; + struct gnttab_unmap_grant_ref unmap[MAX_INDIRECT_SEGMENTS]; + struct page *unmap_pages[MAX_INDIRECT_SEGMENTS]; + struct gntab_unmap_queue_data gnttab_unmap_data; }; diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c index 037f74f..12f9e27 100644 --- a/drivers/net/xen-netback/interface.c +++ b/drivers/net/xen-netback/interface.c @@ -483,9 +483,8 @@ int xenvif_init_queue(struct xenvif_queue *queue) * better enable it. The long term solution would be to use just a * bunch of valid page descriptors, without dependency on ballooning */ - err = alloc_xenballooned_pages(MAX_PENDING_REQS, - queue->mmap_pages, - false); + err = gnttab_alloc_pages(MAX_PENDING_REQS, + queue->mmap_pages); if (err) { netdev_err(queue->vif->dev, "Could not reserve mmap_pages\n"); return -ENOMEM; @@ -664,7 +663,7 @@ void xenvif_disconnect(struct xenvif *vif) */ void xenvif_deinit_queue(struct xenvif_queue *queue) { - free_xenballooned_pages(MAX_PENDING_REQS, queue->mmap_pages); + gnttab_free_pages(MAX_PENDING_REQS, queue->mmap_pages); } void xenvif_free(struct xenvif *vif) diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c index c8ce701..7dc2d64 100644 --- a/drivers/net/xen-netback/netback.c +++ b/drivers/net/xen-netback/netback.c @@ -314,9 +314,7 @@ static struct xenvif_rx_meta *get_next_rx_buffer(struct xenvif_queue *queue, static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb, struct netrx_pending_operations *npo, struct page *page, unsigned long size, - unsigned long offset, int *head, - struct xenvif_queue *foreign_queue, - grant_ref_t foreign_gref) + unsigned long offset, int *head) { struct gnttab_copy *copy_gop; struct xenvif_rx_meta *meta; @@ -333,6 +331,8 @@ static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb offset &= ~PAGE_MASK; while (size > 0) { + struct xen_page_foreign *foreign; + BUG_ON(offset >= PAGE_SIZE); BUG_ON(npo->copy_off > MAX_BUFFER_OFFSET); @@ -361,9 +361,10 @@ static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb copy_gop->flags = GNTCOPY_dest_gref; copy_gop->len = bytes; - if (foreign_queue) { - copy_gop->source.domid = foreign_queue->vif->domid; - copy_gop->source.u.ref = foreign_gref; + foreign = xen_page_foreign(page); + if (foreign) { + copy_gop->source.domid = foreign->domid; + copy_gop->source.u.ref = foreign->gref; copy_gop->flags |= GNTCOPY_source_gref; } else { copy_gop->source.domid = DOMID_SELF; @@ -406,35 +407,6 @@ static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb } /* - * Find the grant ref for a given frag in a chain of struct ubuf_info's - * skb: the skb itself - * i: the frag's number - * ubuf: a pointer to an element in the chain. It should not be NULL - * - * Returns a pointer to the element in the chain where the page were found. If - * not found, returns NULL. - * See the definition of callback_struct in common.h for more details about - * the chain. - */ -static const struct ubuf_info *xenvif_find_gref(const struct sk_buff *const skb, - const int i, - const struct ubuf_info *ubuf) -{ - struct xenvif_queue *foreign_queue = ubuf_to_queue(ubuf); - - do { - u16 pending_idx = ubuf->desc; - - if (skb_shinfo(skb)->frags[i].page.p == - foreign_queue->mmap_pages[pending_idx]) - break; - ubuf = (struct ubuf_info *) ubuf->ctx; - } while (ubuf); - - return ubuf; -} - -/* * Prepare an SKB to be transmitted to the frontend. * * This function is responsible for allocating grant operations, meta @@ -459,8 +431,6 @@ static int xenvif_gop_skb(struct sk_buff *skb, int head = 1; int old_meta_prod; int gso_type; - const struct ubuf_info *ubuf = skb_shinfo(skb)->destructor_arg; - const struct ubuf_info *const head_ubuf = ubuf; old_meta_prod = npo->meta_prod; @@ -507,68 +477,16 @@ static int xenvif_gop_skb(struct sk_buff *skb, len = skb_tail_pointer(skb) - data; xenvif_gop_frag_copy(queue, skb, npo, - virt_to_page(data), len, offset, &head, - NULL, - 0); + virt_to_page(data), len, offset, &head); data += len; } for (i = 0; i < nr_frags; i++) { - /* This variable also signals whether foreign_gref has a real - * value or not. - */ - struct xenvif_queue *foreign_queue = NULL; - grant_ref_t foreign_gref; - - if ((skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY) && - (ubuf->callback == &xenvif_zerocopy_callback)) { - const struct ubuf_info *const startpoint = ubuf; - - /* Ideally ubuf points to the chain element which - * belongs to this frag. Or if frags were removed from - * the beginning, then shortly before it. - */ - ubuf = xenvif_find_gref(skb, i, ubuf); - - /* Try again from the beginning of the list, if we - * haven't tried from there. This only makes sense in - * the unlikely event of reordering the original frags. - * For injected local pages it's an unnecessary second - * run. - */ - if (unlikely(!ubuf) && startpoint != head_ubuf) - ubuf = xenvif_find_gref(skb, i, head_ubuf); - - if (likely(ubuf)) { - u16 pending_idx = ubuf->desc; - - foreign_queue = ubuf_to_queue(ubuf); - foreign_gref = - foreign_queue->pending_tx_info[pending_idx].req.gref; - /* Just a safety measure. If this was the last - * element on the list, the for loop will - * iterate again if a local page were added to - * the end. Using head_ubuf here prevents the - * second search on the chain. Or the original - * frags changed order, but that's less likely. - * In any way, ubuf shouldn't be NULL. - */ - ubuf = ubuf->ctx ? - (struct ubuf_info *) ubuf->ctx : - head_ubuf; - } else - /* This frag was a local page, added to the - * array after the skb left netback. - */ - ubuf = head_ubuf; - } xenvif_gop_frag_copy(queue, skb, npo, skb_frag_page(&skb_shinfo(skb)->frags[i]), skb_frag_size(&skb_shinfo(skb)->frags[i]), skb_shinfo(skb)->frags[i].page_offset, - &head, - foreign_queue, - foreign_queue ? foreign_gref : UINT_MAX); + &head); } return npo->meta_prod - old_meta_prod; @@ -1241,12 +1159,6 @@ static void xenvif_fill_frags(struct xenvif_queue *queue, struct sk_buff *skb) /* Take an extra reference to offset network stack's put_page */ get_page(queue->mmap_pages[pending_idx]); } - /* FIXME: __skb_fill_page_desc set this to true because page->pfmemalloc - * overlaps with "index", and "mapping" is not set. I think mapping - * should be set. If delivered to local stack, it would drop this - * skb in sk_filter unless the socket has the right to use it. - */ - skb->pfmemalloc = false; } static int xenvif_get_extras(struct xenvif_queue *queue, diff --git a/drivers/xen/balloon.c b/drivers/xen/balloon.c index 3860d02..0b52d92 100644 --- a/drivers/xen/balloon.c +++ b/drivers/xen/balloon.c @@ -92,7 +92,6 @@ EXPORT_SYMBOL_GPL(balloon_stats); /* We increase/decrease in batches which fit in a page */ static xen_pfn_t frame_list[PAGE_SIZE / sizeof(unsigned long)]; -static DEFINE_PER_CPU(struct page *, balloon_scratch_page); /* List of ballooned pages, threaded through the mem_map array. */ @@ -423,22 +422,12 @@ static enum bp_state decrease_reservation(unsigned long nr_pages, gfp_t gfp) page = pfn_to_page(pfn); #ifdef CONFIG_XEN_HAVE_PVMMU - /* - * Ballooned out frames are effectively replaced with - * a scratch frame. Ensure direct mappings and the - * p2m are consistent. - */ if (!xen_feature(XENFEAT_auto_translated_physmap)) { if (!PageHighMem(page)) { - struct page *scratch_page = get_balloon_scratch_page(); - ret = HYPERVISOR_update_va_mapping( (unsigned long)__va(pfn << PAGE_SHIFT), - pfn_pte(page_to_pfn(scratch_page), - PAGE_KERNEL_RO), 0); + __pte_ma(0), 0); BUG_ON(ret); - - put_balloon_scratch_page(); } __set_phys_to_machine(pfn, INVALID_P2M_ENTRY); } @@ -500,18 +489,6 @@ static void balloon_process(struct work_struct *work) mutex_unlock(&balloon_mutex); } -struct page *get_balloon_scratch_page(void) -{ - struct page *ret = get_cpu_var(balloon_scratch_page); - BUG_ON(ret == NULL); - return ret; -} - -void put_balloon_scratch_page(void) -{ - put_cpu_var(balloon_scratch_page); -} - /* Resets the Xen limit, sets new target, and kicks off processing. */ void balloon_set_new_target(unsigned long target) { @@ -605,61 +582,13 @@ static void __init balloon_add_region(unsigned long start_pfn, } } -static int alloc_balloon_scratch_page(int cpu) -{ - if (per_cpu(balloon_scratch_page, cpu) != NULL) - return 0; - - per_cpu(balloon_scratch_page, cpu) = alloc_page(GFP_KERNEL); - if (per_cpu(balloon_scratch_page, cpu) == NULL) { - pr_warn("Failed to allocate balloon_scratch_page for cpu %d\n", cpu); - return -ENOMEM; - } - - return 0; -} - - -static int balloon_cpu_notify(struct notifier_block *self, - unsigned long action, void *hcpu) -{ - int cpu = (long)hcpu; - switch (action) { - case CPU_UP_PREPARE: - if (alloc_balloon_scratch_page(cpu)) - return NOTIFY_BAD; - break; - default: - break; - } - return NOTIFY_OK; -} - -static struct notifier_block balloon_cpu_notifier = { - .notifier_call = balloon_cpu_notify, -}; - static int __init balloon_init(void) { - int i, cpu; + int i; if (!xen_domain()) return -ENODEV; - if (!xen_feature(XENFEAT_auto_translated_physmap)) { - register_cpu_notifier(&balloon_cpu_notifier); - - get_online_cpus(); - for_each_online_cpu(cpu) { - if (alloc_balloon_scratch_page(cpu)) { - put_online_cpus(); - unregister_cpu_notifier(&balloon_cpu_notifier); - return -ENOMEM; - } - } - put_online_cpus(); - } - pr_info("Initialising balloon driver\n"); balloon_stats.current_pages = xen_pv_domain() @@ -696,15 +625,4 @@ static int __init balloon_init(void) subsys_initcall(balloon_init); -static int __init balloon_clear(void) -{ - int cpu; - - for_each_possible_cpu(cpu) - per_cpu(balloon_scratch_page, cpu) = NULL; - - return 0; -} -early_initcall(balloon_clear); - MODULE_LICENSE("GPL"); diff --git a/drivers/xen/gntdev.c b/drivers/xen/gntdev.c index 073b4a1..d5bb1a3 100644 --- a/drivers/xen/gntdev.c +++ b/drivers/xen/gntdev.c @@ -67,7 +67,7 @@ struct gntdev_priv { * Only populated if populate_freeable_maps == 1 */ struct list_head freeable_maps; /* lock protects maps and freeable_maps */ - spinlock_t lock; + struct mutex lock; struct mm_struct *mm; struct mmu_notifier mn; }; @@ -91,7 +91,9 @@ struct grant_map { struct gnttab_map_grant_ref *map_ops; struct gnttab_unmap_grant_ref *unmap_ops; struct gnttab_map_grant_ref *kmap_ops; + struct gnttab_unmap_grant_ref *kunmap_ops; struct page **pages; + unsigned long pages_vm_start; }; static int unmap_grant_pages(struct grant_map *map, int offset, int pages); @@ -118,12 +120,13 @@ static void gntdev_free_map(struct grant_map *map) return; if (map->pages) - free_xenballooned_pages(map->count, map->pages); + gnttab_free_pages(map->count, map->pages); kfree(map->pages); kfree(map->grants); kfree(map->map_ops); kfree(map->unmap_ops); kfree(map->kmap_ops); + kfree(map->kunmap_ops); kfree(map); } @@ -140,21 +143,24 @@ static struct grant_map *gntdev_alloc_map(struct gntdev_priv *priv, int count) add->map_ops = kcalloc(count, sizeof(add->map_ops[0]), GFP_KERNEL); add->unmap_ops = kcalloc(count, sizeof(add->unmap_ops[0]), GFP_KERNEL); add->kmap_ops = kcalloc(count, sizeof(add->kmap_ops[0]), GFP_KERNEL); + add->kunmap_ops = kcalloc(count, sizeof(add->kunmap_ops[0]), GFP_KERNEL); add->pages = kcalloc(count, sizeof(add->pages[0]), GFP_KERNEL); if (NULL == add->grants || NULL == add->map_ops || NULL == add->unmap_ops || NULL == add->kmap_ops || + NULL == add->kunmap_ops || NULL == add->pages) goto err; - if (alloc_xenballooned_pages(count, add->pages, false /* lowmem */)) + if (gnttab_alloc_pages(count, add->pages)) goto err; for (i = 0; i < count; i++) { add->map_ops[i].handle = -1; add->unmap_ops[i].handle = -1; add->kmap_ops[i].handle = -1; + add->kunmap_ops[i].handle = -1; } add->index = 0; @@ -216,9 +222,9 @@ static void gntdev_put_map(struct gntdev_priv *priv, struct grant_map *map) } if (populate_freeable_maps && priv) { - spin_lock(&priv->lock); + mutex_lock(&priv->lock); list_del(&map->next); - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); } if (map->pages && !use_ptemod) @@ -239,6 +245,14 @@ static int find_grant_ptes(pte_t *pte, pgtable_t token, BUG_ON(pgnr >= map->count); pte_maddr = arbitrary_virt_to_machine(pte).maddr; + /* + * Set the PTE as special to force get_user_pages_fast() fall + * back to the slow path. If this is not supported as part of + * the grant map, it will be done afterwards. + */ + if (xen_feature(XENFEAT_gnttab_map_avail_bits)) + flags |= (1 << _GNTMAP_guest_avail0); + gnttab_set_map_op(&map->map_ops[pgnr], pte_maddr, flags, map->grants[pgnr].ref, map->grants[pgnr].domid); @@ -247,6 +261,15 @@ static int find_grant_ptes(pte_t *pte, pgtable_t token, return 0; } +#ifdef CONFIG_X86 +static int set_grant_ptes_as_special(pte_t *pte, pgtable_t token, + unsigned long addr, void *data) +{ + set_pte_at(current->mm, addr, pte, pte_mkspecial(*pte)); + return 0; +} +#endif + static int map_grant_pages(struct grant_map *map) { int i, err = 0; @@ -280,6 +303,8 @@ static int map_grant_pages(struct grant_map *map) map->flags | GNTMAP_host_map, map->grants[i].ref, map->grants[i].domid); + gnttab_set_unmap_op(&map->kunmap_ops[i], address, + map->flags | GNTMAP_host_map, -1); } } @@ -290,20 +315,42 @@ static int map_grant_pages(struct grant_map *map) return err; for (i = 0; i < map->count; i++) { - if (map->map_ops[i].status) + if (map->map_ops[i].status) { err = -EINVAL; - else { - BUG_ON(map->map_ops[i].handle == -1); - map->unmap_ops[i].handle = map->map_ops[i].handle; - pr_debug("map handle=%d\n", map->map_ops[i].handle); + continue; } + + map->unmap_ops[i].handle = map->map_ops[i].handle; + if (use_ptemod) + map->kunmap_ops[i].handle = map->kmap_ops[i].handle; } return err; } +struct unmap_grant_pages_callback_data +{ + struct completion completion; + int result; +}; + +static void unmap_grant_callback(int result, + struct gntab_unmap_queue_data *data) +{ + struct unmap_grant_pages_callback_data* d = data->data; + + d->result = result; + complete(&d->completion); +} + static int __unmap_grant_pages(struct grant_map *map, int offset, int pages) { int i, err = 0; + struct gntab_unmap_queue_data unmap_data; + struct unmap_grant_pages_callback_data data; + + init_completion(&data.completion); + unmap_data.data = &data; + unmap_data.done= &unmap_grant_callback; if (map->notify.flags & UNMAP_NOTIFY_CLEAR_BYTE) { int pgno = (map->notify.addr >> PAGE_SHIFT); @@ -315,11 +362,16 @@ static int __unmap_grant_pages(struct grant_map *map, int offset, int pages) } } - err = gnttab_unmap_refs(map->unmap_ops + offset, - use_ptemod ? map->kmap_ops + offset : NULL, map->pages + offset, - pages); - if (err) - return err; + unmap_data.unmap_ops = map->unmap_ops + offset; + unmap_data.kunmap_ops = use_ptemod ? map->kunmap_ops + offset : NULL; + unmap_data.pages = map->pages + offset; + unmap_data.count = pages; + + gnttab_unmap_refs_async(&unmap_data); + + wait_for_completion(&data.completion); + if (data.result) + return data.result; for (i = 0; i < pages; i++) { if (map->unmap_ops[offset+i].status) @@ -387,17 +439,26 @@ static void gntdev_vma_close(struct vm_area_struct *vma) * not do any unmapping, since that has been done prior to * closing the vma, but it may still iterate the unmap_ops list. */ - spin_lock(&priv->lock); + mutex_lock(&priv->lock); map->vma = NULL; - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); } vma->vm_private_data = NULL; gntdev_put_map(priv, map); } +static struct page *gntdev_vma_find_special_page(struct vm_area_struct *vma, + unsigned long addr) +{ + struct grant_map *map = vma->vm_private_data; + + return map->pages[(addr - map->pages_vm_start) >> PAGE_SHIFT]; +} + static struct vm_operations_struct gntdev_vmops = { .open = gntdev_vma_open, .close = gntdev_vma_close, + .find_special_page = gntdev_vma_find_special_page, }; /* ------------------------------------------------------------------ */ @@ -433,14 +494,14 @@ static void mn_invl_range_start(struct mmu_notifier *mn, struct gntdev_priv *priv = container_of(mn, struct gntdev_priv, mn); struct grant_map *map; - spin_lock(&priv->lock); + mutex_lock(&priv->lock); list_for_each_entry(map, &priv->maps, next) { unmap_if_in_range(map, start, end); } list_for_each_entry(map, &priv->freeable_maps, next) { unmap_if_in_range(map, start, end); } - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); } static void mn_invl_page(struct mmu_notifier *mn, @@ -457,7 +518,7 @@ static void mn_release(struct mmu_notifier *mn, struct grant_map *map; int err; - spin_lock(&priv->lock); + mutex_lock(&priv->lock); list_for_each_entry(map, &priv->maps, next) { if (!map->vma) continue; @@ -476,7 +537,7 @@ static void mn_release(struct mmu_notifier *mn, err = unmap_grant_pages(map, /* offset */ 0, map->count); WARN_ON(err); } - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); } static struct mmu_notifier_ops gntdev_mmu_ops = { @@ -498,7 +559,7 @@ static int gntdev_open(struct inode *inode, struct file *flip) INIT_LIST_HEAD(&priv->maps); INIT_LIST_HEAD(&priv->freeable_maps); - spin_lock_init(&priv->lock); + mutex_init(&priv->lock); if (use_ptemod) { priv->mm = get_task_mm(current); @@ -572,10 +633,10 @@ static long gntdev_ioctl_map_grant_ref(struct gntdev_priv *priv, return -EFAULT; } - spin_lock(&priv->lock); + mutex_lock(&priv->lock); gntdev_add_map(priv, map); op.index = map->index << PAGE_SHIFT; - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); if (copy_to_user(u, &op, sizeof(op)) != 0) return -EFAULT; @@ -594,7 +655,7 @@ static long gntdev_ioctl_unmap_grant_ref(struct gntdev_priv *priv, return -EFAULT; pr_debug("priv %p, del %d+%d\n", priv, (int)op.index, (int)op.count); - spin_lock(&priv->lock); + mutex_lock(&priv->lock); map = gntdev_find_map_index(priv, op.index >> PAGE_SHIFT, op.count); if (map) { list_del(&map->next); @@ -602,7 +663,7 @@ static long gntdev_ioctl_unmap_grant_ref(struct gntdev_priv *priv, list_add_tail(&map->next, &priv->freeable_maps); err = 0; } - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); if (map) gntdev_put_map(priv, map); return err; @@ -670,7 +731,7 @@ static long gntdev_ioctl_notify(struct gntdev_priv *priv, void __user *u) out_flags = op.action; out_event = op.event_channel_port; - spin_lock(&priv->lock); + mutex_lock(&priv->lock); list_for_each_entry(map, &priv->maps, next) { uint64_t begin = map->index << PAGE_SHIFT; @@ -698,7 +759,7 @@ static long gntdev_ioctl_notify(struct gntdev_priv *priv, void __user *u) rc = 0; unlock_out: - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); /* Drop the reference to the event channel we did not save in the map */ if (out_flags & UNMAP_NOTIFY_SEND_EVENT) @@ -748,7 +809,7 @@ static int gntdev_mmap(struct file *flip, struct vm_area_struct *vma) pr_debug("map %d+%d at %lx (pgoff %lx)\n", index, count, vma->vm_start, vma->vm_pgoff); - spin_lock(&priv->lock); + mutex_lock(&priv->lock); map = gntdev_find_map_index(priv, index, count); if (!map) goto unlock_out; @@ -783,7 +844,7 @@ static int gntdev_mmap(struct file *flip, struct vm_area_struct *vma) map->flags |= GNTMAP_readonly; } - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); if (use_ptemod) { err = apply_to_page_range(vma->vm_mm, vma->vm_start, @@ -806,16 +867,34 @@ static int gntdev_mmap(struct file *flip, struct vm_area_struct *vma) if (err) goto out_put_map; } + } else { +#ifdef CONFIG_X86 + /* + * If the PTEs were not made special by the grant map + * hypercall, do so here. + * + * This is racy since the mapping is already visible + * to userspace but userspace should be well-behaved + * enough to not touch it until the mmap() call + * returns. + */ + if (!xen_feature(XENFEAT_gnttab_map_avail_bits)) { + apply_to_page_range(vma->vm_mm, vma->vm_start, + vma->vm_end - vma->vm_start, + set_grant_ptes_as_special, NULL); + } +#endif + map->pages_vm_start = vma->vm_start; } return 0; unlock_out: - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); return err; out_unlock_put: - spin_unlock(&priv->lock); + mutex_unlock(&priv->lock); out_put_map: if (use_ptemod) map->vma = NULL; diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c index 7786291..17972fb 100644 --- a/drivers/xen/grant-table.c +++ b/drivers/xen/grant-table.c @@ -42,6 +42,7 @@ #include <linux/io.h> #include <linux/delay.h> #include <linux/hardirq.h> +#include <linux/workqueue.h> #include <xen/xen.h> #include <xen/interface/xen.h> @@ -50,6 +51,7 @@ #include <xen/interface/memory.h> #include <xen/hvc-console.h> #include <xen/swiotlb-xen.h> +#include <xen/balloon.h> #include <asm/xen/hypercall.h> #include <asm/xen/interface.h> @@ -671,6 +673,59 @@ void gnttab_free_auto_xlat_frames(void) } EXPORT_SYMBOL_GPL(gnttab_free_auto_xlat_frames); +/** + * gnttab_alloc_pages - alloc pages suitable for grant mapping into + * @nr_pages: number of pages to alloc + * @pages: returns the pages + */ +int gnttab_alloc_pages(int nr_pages, struct page **pages) +{ + int i; + int ret; + + ret = alloc_xenballooned_pages(nr_pages, pages, false); + if (ret < 0) + return ret; + + for (i = 0; i < nr_pages; i++) { +#if BITS_PER_LONG < 64 + struct xen_page_foreign *foreign; + + foreign = kzalloc(sizeof(*foreign), GFP_KERNEL); + if (!foreign) { + gnttab_free_pages(nr_pages, pages); + return -ENOMEM; + } + set_page_private(pages[i], (unsigned long)foreign); +#endif + SetPagePrivate(pages[i]); + } + + return 0; +} +EXPORT_SYMBOL(gnttab_alloc_pages); + +/** + * gnttab_free_pages - free pages allocated by gnttab_alloc_pages() + * @nr_pages; number of pages to free + * @pages: the pages + */ +void gnttab_free_pages(int nr_pages, struct page **pages) +{ + int i; + + for (i = 0; i < nr_pages; i++) { + if (PagePrivate(pages[i])) { +#if BITS_PER_LONG < 64 + kfree((void *)page_private(pages[i])); +#endif + ClearPagePrivate(pages[i]); + } + } + free_xenballooned_pages(nr_pages, pages); +} +EXPORT_SYMBOL(gnttab_free_pages); + /* Handling of paged out grant targets (GNTST_eagain) */ #define MAX_DELAY 256 static inline void @@ -727,30 +782,87 @@ int gnttab_map_refs(struct gnttab_map_grant_ref *map_ops, if (ret) return ret; - /* Retry eagain maps */ - for (i = 0; i < count; i++) + for (i = 0; i < count; i++) { + /* Retry eagain maps */ if (map_ops[i].status == GNTST_eagain) gnttab_retry_eagain_gop(GNTTABOP_map_grant_ref, map_ops + i, &map_ops[i].status, __func__); + if (map_ops[i].status == GNTST_okay) { + struct xen_page_foreign *foreign; + + SetPageForeign(pages[i]); + foreign = xen_page_foreign(pages[i]); + foreign->domid = map_ops[i].dom; + foreign->gref = map_ops[i].ref; + } + } + return set_foreign_p2m_mapping(map_ops, kmap_ops, pages, count); } EXPORT_SYMBOL_GPL(gnttab_map_refs); int gnttab_unmap_refs(struct gnttab_unmap_grant_ref *unmap_ops, - struct gnttab_map_grant_ref *kmap_ops, + struct gnttab_unmap_grant_ref *kunmap_ops, struct page **pages, unsigned int count) { + unsigned int i; int ret; ret = HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, unmap_ops, count); if (ret) return ret; - return clear_foreign_p2m_mapping(unmap_ops, kmap_ops, pages, count); + for (i = 0; i < count; i++) + ClearPageForeign(pages[i]); + + return clear_foreign_p2m_mapping(unmap_ops, kunmap_ops, pages, count); } EXPORT_SYMBOL_GPL(gnttab_unmap_refs); +#define GNTTAB_UNMAP_REFS_DELAY 5 + +static void __gnttab_unmap_refs_async(struct gntab_unmap_queue_data* item); + +static void gnttab_unmap_work(struct work_struct *work) +{ + struct gntab_unmap_queue_data + *unmap_data = container_of(work, + struct gntab_unmap_queue_data, + gnttab_work.work); + if (unmap_data->age != UINT_MAX) + unmap_data->age++; + __gnttab_unmap_refs_async(unmap_data); +} + +static void __gnttab_unmap_refs_async(struct gntab_unmap_queue_data* item) +{ + int ret; + int pc; + + for (pc = 0; pc < item->count; pc++) { + if (page_count(item->pages[pc]) > 1) { + unsigned long delay = GNTTAB_UNMAP_REFS_DELAY * (item->age + 1); + schedule_delayed_work(&item->gnttab_work, + msecs_to_jiffies(delay)); + return; + } + } + + ret = gnttab_unmap_refs(item->unmap_ops, item->kunmap_ops, + item->pages, item->count); + item->done(ret, item); +} + +void gnttab_unmap_refs_async(struct gntab_unmap_queue_data* item) +{ + INIT_DELAYED_WORK(&item->gnttab_work, gnttab_unmap_work); + item->age = 0; + + __gnttab_unmap_refs_async(item); +} +EXPORT_SYMBOL_GPL(gnttab_unmap_refs_async); + static int gnttab_map_frames_v1(xen_pfn_t *frames, unsigned int nr_gframes) { int rc; diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c index f8bb36f..bf19407 100644 --- a/drivers/xen/manage.c +++ b/drivers/xen/manage.c @@ -105,10 +105,16 @@ static void do_suspend(void) err = freeze_processes(); if (err) { - pr_err("%s: freeze failed %d\n", __func__, err); + pr_err("%s: freeze processes failed %d\n", __func__, err); goto out; } + err = freeze_kernel_threads(); + if (err) { + pr_err("%s: freeze kernel threads failed %d\n", __func__, err); + goto out_thaw; + } + err = dpm_suspend_start(PMSG_FREEZE); if (err) { pr_err("%s: dpm_suspend_start %d\n", __func__, err); diff --git a/drivers/xen/tmem.c b/drivers/xen/tmem.c index 83b5c53..8a65423 100644 --- a/drivers/xen/tmem.c +++ b/drivers/xen/tmem.c @@ -374,7 +374,7 @@ static struct frontswap_ops tmem_frontswap_ops = { }; #endif -static int xen_tmem_init(void) +static int __init xen_tmem_init(void) { if (!xen_domain()) return 0; diff --git a/drivers/xen/xen-scsiback.c b/drivers/xen/xen-scsiback.c index e999496e..ecd540a 100644 --- a/drivers/xen/xen-scsiback.c +++ b/drivers/xen/xen-scsiback.c @@ -227,7 +227,7 @@ static void put_free_pages(struct page **page, int num) return; if (i > scsiback_max_buffer_pages) { n = min(num, i - scsiback_max_buffer_pages); - free_xenballooned_pages(n, page + num - n); + gnttab_free_pages(n, page + num - n); n = num - n; } spin_lock_irqsave(&free_pages_lock, flags); @@ -244,7 +244,7 @@ static int get_free_page(struct page **page) spin_lock_irqsave(&free_pages_lock, flags); if (list_empty(&scsiback_free_pages)) { spin_unlock_irqrestore(&free_pages_lock, flags); - return alloc_xenballooned_pages(1, page, false); + return gnttab_alloc_pages(1, page); } page[0] = list_first_entry(&scsiback_free_pages, struct page, lru); list_del(&page[0]->lru); @@ -2106,7 +2106,7 @@ static void __exit scsiback_exit(void) while (free_pages_num) { if (get_free_page(&page)) BUG(); - free_xenballooned_pages(1, &page); + gnttab_free_pages(1, &page); } scsiback_deregister_configfs(); xenbus_unregister_driver(&scsiback_driver); diff --git a/drivers/xen/xenbus/xenbus_dev_frontend.c b/drivers/xen/xenbus/xenbus_dev_frontend.c index 85534ea..9433e46 100644 --- a/drivers/xen/xenbus/xenbus_dev_frontend.c +++ b/drivers/xen/xenbus/xenbus_dev_frontend.c @@ -326,10 +326,13 @@ static int xenbus_write_transaction(unsigned msg_type, } if (msg_type == XS_TRANSACTION_START) { - trans->handle.id = simple_strtoul(reply, NULL, 0); - - list_add(&trans->list, &u->transactions); - } else if (msg_type == XS_TRANSACTION_END) { + if (u->u.msg.type == XS_ERROR) + kfree(trans); + else { + trans->handle.id = simple_strtoul(reply, NULL, 0); + list_add(&trans->list, &u->transactions); + } + } else if (u->u.msg.type == XS_TRANSACTION_END) { list_for_each_entry(trans, &u->transactions, list) if (trans->handle.id == u->u.msg.tx_id) break; |