- Linux-stable-mirror - lists.linaro.org

FAILED: patch "[PATCH] fuse: continue to send FUSE_RELEASEDIR when FUSE_OPEN returns" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 2e64ff154ce6ce9a8dc0f9556463916efa6ff460 Mon Sep 17 00:00:00 2001 From: Chad Austin <chadaustin(a)fb.com> Date: Mon, 10 Dec 2018 10:54:52 -0800 Subject: [PATCH] fuse: continue to send FUSE_RELEASEDIR when FUSE_OPEN returns ENOSYS When FUSE_OPEN returns ENOSYS, the no_open bit is set on the connection. Because the FUSE_RELEASE and FUSE_RELEASEDIR paths share code, this incorrectly caused the FUSE_RELEASEDIR request to be dropped and never sent to userspace. Pass an isdir bool to distinguish between FUSE_RELEASE and FUSE_RELEASEDIR inside of fuse_file_put. Fixes: 7678ac50615d ("fuse: support clients that don't implement 'open'") Cc: <stable(a)vger.kernel.org> # v3.14 Signed-off-by: Chad Austin <chadaustin(a)fb.com> Signed-off-by: Miklos Szeredi <mszeredi(a)redhat.com> diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index dc4e83d8ace7..e909678afa2d 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1243,7 +1243,7 @@ static int fuse_dir_open(struct inode *inode, struct file *file) static int fuse_dir_release(struct inode *inode, struct file *file) { - fuse_release_common(file, FUSE_RELEASEDIR); + fuse_release_common(file, true); return 0; } diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 677c51341e96..ffaffe18352a 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -89,12 +89,12 @@ static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req) iput(req->misc.release.inode); } -static void fuse_file_put(struct fuse_file *ff, bool sync) +static void fuse_file_put(struct fuse_file *ff, bool sync, bool isdir) { if (refcount_dec_and_test(&ff->count)) { struct fuse_req *req = ff->reserved_req; - if (ff->fc->no_open) { + if (ff->fc->no_open && !isdir) { /* * Drop the release request when client does not * implement 'open' @@ -247,10 +247,11 @@ static void fuse_prepare_release(struct fuse_file *ff, int flags, int opcode) req->in.args[0].value = inarg; } -void fuse_release_common(struct file *file, int opcode) +void fuse_release_common(struct file *file, bool isdir) { struct fuse_file *ff = file->private_data; struct fuse_req *req = ff->reserved_req; + int opcode = isdir ? FUSE_RELEASEDIR : FUSE_RELEASE; fuse_prepare_release(ff, file->f_flags, opcode); @@ -272,7 +273,7 @@ void fuse_release_common(struct file *file, int opcode) * synchronous RELEASE is allowed (and desirable) in this case * because the server can be trusted not to screw up. */ - fuse_file_put(ff, ff->fc->destroy_req != NULL); + fuse_file_put(ff, ff->fc->destroy_req != NULL, isdir); } static int fuse_open(struct inode *inode, struct file *file) @@ -288,7 +289,7 @@ static int fuse_release(struct inode *inode, struct file *file) if (fc->writeback_cache) write_inode_now(inode, 1); - fuse_release_common(file, FUSE_RELEASE); + fuse_release_common(file, false); /* return value is ignored by VFS */ return 0; @@ -302,7 +303,7 @@ void fuse_sync_release(struct fuse_file *ff, int flags) * iput(NULL) is a no-op and since the refcount is 1 and everything's * synchronous, we are fine with not doing igrab() here" */ - fuse_file_put(ff, true); + fuse_file_put(ff, true, false); } EXPORT_SYMBOL_GPL(fuse_sync_release); @@ -808,7 +809,7 @@ static void fuse_readpages_end(struct fuse_conn *fc, struct fuse_req *req) put_page(page); } if (req->ff) - fuse_file_put(req->ff, false); + fuse_file_put(req->ff, false, false); } static void fuse_send_readpages(struct fuse_req *req, struct file *file) @@ -1461,7 +1462,7 @@ static void fuse_writepage_free(struct fuse_conn *fc, struct fuse_req *req) __free_page(req->pages[i]); if (req->ff) - fuse_file_put(req->ff, false); + fuse_file_put(req->ff, false, false); } static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req) @@ -1620,7 +1621,7 @@ int fuse_write_inode(struct inode *inode, struct writeback_control *wbc) ff = __fuse_write_file_get(fc, fi); err = fuse_flush_times(inode, ff); if (ff) - fuse_file_put(ff, 0); + fuse_file_put(ff, false, false); return err; } @@ -1941,7 +1942,7 @@ static int fuse_writepages(struct address_space *mapping, err = 0; } if (data.ff) - fuse_file_put(data.ff, false); + fuse_file_put(data.ff, false, false); kfree(data.orig_pages); out: diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index afe1f231c758..2f2c92e6f8cb 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -822,7 +822,7 @@ void fuse_sync_release(struct fuse_file *ff, int flags); /** * Send RELEASE or RELEASEDIR request */ -void fuse_release_common(struct file *file, int opcode); +void fuse_release_common(struct file *file, bool isdir); /** * Send FSYNC or FSYNCDIR request

6 years, 9 months

2
1
0 0

[PATCH] IB/hfi1: Remove race conditions in user_sdma send path

by Michael J. Ruhl

From: Michael J. Ruhl <michael.j.ruhl(a)intel.com> commit 28a9a9e83ceae2cee25b9af9ad20d53aaa9ab951 upstream Packet queue state is over used to determine SDMA descriptor availablitity and packet queue request state. cpu 0 ret = user_sdma_send_pkts(req, pcount); cpu 0 if (atomic_read(&pq->n_reqs)) cpu 1 IRQ user_sdma_txreq_cb calls pq_update() (state to _INACTIVE) cpu 0 xchg(&pq->state, SDMA_PKT_Q_ACTIVE); At this point pq->n_reqs == 0 and pq->state is incorrectly SDMA_PKT_Q_ACTIVE. The close path will hang waiting for the state to return to _INACTIVE. This can also change the state from _DEFERRED to _ACTIVE. However, this is a mostly benign race. Remove the racy code path. Use n_reqs to determine if a packet queue is active or not. Cc: <stable(a)vger.kernel.org> # 4.19.x Reviewed-by: Mitko Haralanov <mitko.haralanov(a)intel.com> Reviewed-by: Mike Marciniszyn <mike.marciniszyn(a)intel.com> Signed-off-by: Michael J. Ruhl <michael.j.ruhl(a)intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro(a)intel.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> --- drivers/infiniband/hw/hfi1/user_sdma.c | 24 ++++++++++-------------- drivers/infiniband/hw/hfi1/user_sdma.h | 9 +++++---- 2 files changed, 15 insertions(+), 18 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/user_sdma.c b/drivers/infiniband/hw/hfi1/user_sdma.c index 39134dd..51831bf 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.c +++ b/drivers/infiniband/hw/hfi1/user_sdma.c @@ -187,7 +187,6 @@ int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, pq->ctxt = uctxt->ctxt; pq->subctxt = fd->subctxt; pq->n_max_reqs = hfi1_sdma_comp_ring_size; - pq->state = SDMA_PKT_Q_INACTIVE; atomic_set(&pq->n_reqs, 0); init_waitqueue_head(&pq->wait); atomic_set(&pq->n_locked, 0); @@ -276,7 +275,7 @@ int hfi1_user_sdma_free_queues(struct hfi1_filedata *fd, /* Wait until all requests have been freed. */ wait_event_interruptible( pq->wait, - (READ_ONCE(pq->state) == SDMA_PKT_Q_INACTIVE)); + !atomic_read(&pq->n_reqs)); kfree(pq->reqs); kfree(pq->req_in_use); kmem_cache_destroy(pq->txreq_cache); @@ -312,6 +311,13 @@ static u8 dlid_to_selector(u16 dlid) return mapping[hash]; } +/** + * hfi1_user_sdma_process_request() - Process and start a user sdma request + * @fd: valid file descriptor + * @iovec: array of io vectors to process + * @dim: overall iovec array size + * @count: number of io vector array entries processed + */ int hfi1_user_sdma_process_request(struct hfi1_filedata *fd, struct iovec *iovec, unsigned long dim, unsigned long *count) @@ -560,21 +566,13 @@ int hfi1_user_sdma_process_request(struct hfi1_filedata *fd, req->ahg_idx = sdma_ahg_alloc(req->sde); set_comp_state(pq, cq, info.comp_idx, QUEUED, 0); + pq->state = SDMA_PKT_Q_ACTIVE; /* Send the first N packets in the request to buy us some time */ ret = user_sdma_send_pkts(req, pcount); if (unlikely(ret < 0 && ret != -EBUSY)) goto free_req; /* - * It is possible that the SDMA engine would have processed all the - * submitted packets by the time we get here. Therefore, only set - * packet queue state to ACTIVE if there are still uncompleted - * requests. - */ - if (atomic_read(&pq->n_reqs)) - xchg(&pq->state, SDMA_PKT_Q_ACTIVE); - - /* * This is a somewhat blocking send implementation. * The driver will block the caller until all packets of the * request have been submitted to the SDMA engine. However, it @@ -1409,10 +1407,8 @@ static void user_sdma_txreq_cb(struct sdma_txreq *txreq, int status) static inline void pq_update(struct hfi1_user_sdma_pkt_q *pq) { - if (atomic_dec_and_test(&pq->n_reqs)) { - xchg(&pq->state, SDMA_PKT_Q_INACTIVE); + if (atomic_dec_and_test(&pq->n_reqs)) wake_up(&pq->wait); - } } static void user_sdma_free_request(struct user_sdma_request *req, bool unpin) diff --git a/drivers/infiniband/hw/hfi1/user_sdma.h b/drivers/infiniband/hw/hfi1/user_sdma.h index 0ae0645..91c343f 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.h +++ b/drivers/infiniband/hw/hfi1/user_sdma.h @@ -105,9 +105,10 @@ static inline int ahg_header_set(u32 *arr, int idx, size_t array_size, #define TXREQ_FLAGS_REQ_ACK BIT(0) /* Set the ACK bit in the header */ #define TXREQ_FLAGS_REQ_DISABLE_SH BIT(1) /* Disable header suppression */ -#define SDMA_PKT_Q_INACTIVE BIT(0) -#define SDMA_PKT_Q_ACTIVE BIT(1) -#define SDMA_PKT_Q_DEFERRED BIT(2) +enum pkt_q_sdma_state { + SDMA_PKT_Q_ACTIVE, + SDMA_PKT_Q_DEFERRED, +}; /* * Maximum retry attempts to submit a TX request @@ -133,7 +134,7 @@ struct hfi1_user_sdma_pkt_q { struct user_sdma_request *reqs; unsigned long *req_in_use; struct iowait busy; - unsigned state; + enum pkt_q_sdma_state state; wait_queue_head_t wait; unsigned long unpinned; struct mmu_rb_handler *handler;

6 years, 9 months

2
1
0 0

[PATCH] dm: give req_size precisely given by user

by Jaegeuk Kim

This patch makes dm be aware of io_pages to assign sane req_size for reads. Cc: stable(a)vger.kernel.org Cc: Jens Axboe <axboe(a)kernel.dk> Cc: Alasdair Kergon <agk(a)redhat.com> Cc: Mike Snitzer <snitzer(a)redhat.com> Cc: dm-devel(a)redhat.com Fixes: 9491ae4a ("mm: don't cap request size based on read-ahead setting") Signed-off-by: Jaegeuk Kim <jaegeuk(a)kernel.org> --- drivers/md/dm-table.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c index 9038c302d5c2..886790428241 100644 --- a/drivers/md/dm-table.c +++ b/drivers/md/dm-table.c @@ -1927,6 +1927,9 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q, */ if (blk_queue_is_zoned(q)) blk_revalidate_disk_zones(t->md->disk); + + /* io_pages is used for readahead */ + q->backing_dev_info->io_pages = limits->max_sectors >> (PAGE_SHIFT - 9); } unsigned int dm_table_get_num_targets(struct dm_table *t) -- 2.19.0.605.g01d371f741-goog

6 years, 9 months

3
3
0 0

[PATCH 12/25] clocksource/drivers/arc_timer: Utilize generic sched_clock

by Daniel Lezcano

From: Alexey Brodkin <alexey.brodkin(a)synopsys.com> It turned out we used to use default implementation of sched_clock() from kernel/sched/clock.c which was as precise as 1/HZ, i.e. by default we had 10 msec granularity of time measurement. Now given ARC built-in timers are clocked with the same frequency as CPU cores we may get much higher precision of time tracking. Thus we switch to generic sched_clock which really reads ARC hardware counters. This is especially helpful for measuring short events. That's what we used to have: ------------------------------>8------------------------ $ perf stat /bin/sh -c /root/lmbench-master/bin/arc/hello > /dev/null Performance counter stats for '/bin/sh -c /root/lmbench-master/bin/arc/hello': 10.000000 task-clock (msec) # 2.832 CPUs utilized 1 context-switches # 0.100 K/sec 1 cpu-migrations # 0.100 K/sec 63 page-faults # 0.006 M/sec 3049480 cycles # 0.305 GHz 1091259 instructions # 0.36 insn per cycle 256828 branches # 25.683 M/sec 27026 branch-misses # 10.52% of all branches 0.003530687 seconds time elapsed 0.000000000 seconds user 0.010000000 seconds sys ------------------------------>8------------------------ And now we'll see: ------------------------------>8------------------------ $ perf stat /bin/sh -c /root/lmbench-master/bin/arc/hello > /dev/null Performance counter stats for '/bin/sh -c /root/lmbench-master/bin/arc/hello': 3.004322 task-clock (msec) # 0.865 CPUs utilized 1 context-switches # 0.333 K/sec 1 cpu-migrations # 0.333 K/sec 63 page-faults # 0.021 M/sec 2986734 cycles # 0.994 GHz 1087466 instructions # 0.36 insn per cycle 255209 branches # 84.947 M/sec 26002 branch-misses # 10.19% of all branches 0.003474829 seconds time elapsed 0.003519000 seconds user 0.000000000 seconds sys ------------------------------>8------------------------ Note how much more meaningful is the second output - time spent for execution pretty much matches number of cycles spent (we're runnign @ 1GHz here). Signed-off-by: Alexey Brodkin <abrodkin(a)synopsys.com> Cc: Daniel Lezcano <daniel.lezcano(a)linaro.org> Cc: Vineet Gupta <vgupta(a)synopsys.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: stable(a)vger.kernel.org Acked-by: Vineet Gupta <vgupta(a)synopsys.com> Signed-off-by: Daniel Lezcano <daniel.lezcano(a)linaro.org> Signed-off-by: Daniel Lezcano <daniel.lezcano(a)linaro.org> --- arch/arc/Kconfig | 1 + drivers/clocksource/Kconfig | 1 + drivers/clocksource/arc_timer.c | 22 ++++++++++++++++++++++ 3 files changed, 24 insertions(+) diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig index c9e2a1323536..74b5a654f664 100644 --- a/arch/arc/Kconfig +++ b/arch/arc/Kconfig @@ -26,6 +26,7 @@ config ARC select GENERIC_IRQ_SHOW select GENERIC_PCI_IOMAP select GENERIC_PENDING_IRQ if SMP + select GENERIC_SCHED_CLOCK select GENERIC_SMP_IDLE_THREAD select HAVE_ARCH_KGDB select HAVE_ARCH_TRACEHOOK diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig index 8761a1c21b6c..c57b156f49a2 100644 --- a/drivers/clocksource/Kconfig +++ b/drivers/clocksource/Kconfig @@ -277,6 +277,7 @@ config CLKSRC_MPS2 config ARC_TIMERS bool "Support for 32-bit TIMERn counters in ARC Cores" if COMPILE_TEST + depends on GENERIC_SCHED_CLOCK select TIMER_OF help These are legacy 32-bit TIMER0 and TIMER1 counters found on all ARC cores diff --git a/drivers/clocksource/arc_timer.c b/drivers/clocksource/arc_timer.c index 20da9b1d7f7d..b28970ca4a7a 100644 --- a/drivers/clocksource/arc_timer.c +++ b/drivers/clocksource/arc_timer.c @@ -23,6 +23,7 @@ #include <linux/cpu.h> #include <linux/of.h> #include <linux/of_irq.h> +#include <linux/sched_clock.h> #include <soc/arc/timers.h> #include <soc/arc/mcip.h> @@ -88,6 +89,11 @@ static u64 arc_read_gfrc(struct clocksource *cs) return (((u64)h) << 32) | l; } +static notrace u64 arc_gfrc_clock_read(void) +{ + return arc_read_gfrc(NULL); +} + static struct clocksource arc_counter_gfrc = { .name = "ARConnect GFRC", .rating = 400, @@ -111,6 +117,8 @@ static int __init arc_cs_setup_gfrc(struct device_node *node) if (ret) return ret; + sched_clock_register(arc_gfrc_clock_read, 64, arc_timer_freq); + return clocksource_register_hz(&arc_counter_gfrc, arc_timer_freq); } TIMER_OF_DECLARE(arc_gfrc, "snps,archs-timer-gfrc", arc_cs_setup_gfrc); @@ -139,6 +147,11 @@ static u64 arc_read_rtc(struct clocksource *cs) return (((u64)h) << 32) | l; } +static notrace u64 arc_rtc_clock_read(void) +{ + return arc_read_rtc(NULL); +} + static struct clocksource arc_counter_rtc = { .name = "ARCv2 RTC", .rating = 350, @@ -170,6 +183,8 @@ static int __init arc_cs_setup_rtc(struct device_node *node) write_aux_reg(AUX_RTC_CTRL, 1); + sched_clock_register(arc_rtc_clock_read, 64, arc_timer_freq); + return clocksource_register_hz(&arc_counter_rtc, arc_timer_freq); } TIMER_OF_DECLARE(arc_rtc, "snps,archs-timer-rtc", arc_cs_setup_rtc); @@ -185,6 +200,11 @@ static u64 arc_read_timer1(struct clocksource *cs) return (u64) read_aux_reg(ARC_REG_TIMER1_CNT); } +static notrace u64 arc_timer1_clock_read(void) +{ + return arc_read_timer1(NULL); +} + static struct clocksource arc_counter_timer1 = { .name = "ARC Timer1", .rating = 300, @@ -209,6 +229,8 @@ static int __init arc_cs_setup_timer1(struct device_node *node) write_aux_reg(ARC_REG_TIMER1_CNT, 0); write_aux_reg(ARC_REG_TIMER1_CTRL, TIMER_CTRL_NH); + sched_clock_register(arc_timer1_clock_read, 32, arc_timer_freq); + return clocksource_register_hz(&arc_counter_timer1, arc_timer_freq); } -- 2.17.1

6 years, 9 months

1
0
0 0

stable-rc/linux-4.19.y boot: 117 boots: 0 failed, 104 passed with 13 offline (v4.19.10-45-g75803dcafdb3)

by kernelci.org bot

stable-rc/linux-4.19.y boot: 117 boots: 0 failed, 104 passed with 13 offline (v4.19.10-45-g75803dcafdb3) Full Boot Summary: https://kernelci.org/boot/all/job/stable-rc/branch/linux-4.19.y/kernel/v4.1… Full Build Summary: https://kernelci.org/build/stable-rc/branch/linux-4.19.y/kernel/v4.19.10-45… Tree: stable-rc Branch: linux-4.19.y Git Describe: v4.19.10-45-g75803dcafdb3 Git Commit: 75803dcafdb3171314d8767b002143fc031611e9 Git URL: http://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git Tested: 63 unique boards, 23 SoC families, 12 builds out of 202 Offline Platforms: arm: tegra_defconfig: tegra124-jetson-tk1: 1 offline lab multi_v7_defconfig: alpine-db: 1 offline lab at91-sama5d4_xplained: 1 offline lab socfpga_cyclone5_de0_sockit: 1 offline lab stih410-b2120: 1 offline lab sun5i-r8-chip: 1 offline lab tegra124-jetson-tk1: 1 offline lab bcm2835_defconfig: bcm2835-rpi-b: 1 offline lab sunxi_defconfig: sun5i-r8-chip: 1 offline lab sama5_defconfig: at91-sama5d4_xplained: 1 offline lab arm64: defconfig: apq8016-sbc: 1 offline lab juno-r2: 1 offline lab mt7622-rfb1: 1 offline lab --- For more info write to <info(a)kernelci.org>

6 years, 9 months

1
0
0 0

[PATCH] IB/hfi1: Remove race conditions in user_sdma send path

by Michael J. Ruhl

From: Michael J. Ruhl <michael.j.ruhl(a)intel.com> commit 28a9a9e83ceae2cee25b9af9ad20d53aaa9ab951 upstream Packet queue state is over used to determine SDMA descriptor availablitity and packet queue request state. cpu 0 ret = user_sdma_send_pkts(req, pcount); cpu 0 if (atomic_read(&pq->n_reqs)) cpu 1 IRQ user_sdma_txreq_cb calls pq_update() (state to _INACTIVE) cpu 0 xchg(&pq->state, SDMA_PKT_Q_ACTIVE); At this point pq->n_reqs == 0 and pq->state is incorrectly SDMA_PKT_Q_ACTIVE. The close path will hang waiting for the state to return to _INACTIVE. This can also change the state from _DEFERRED to _ACTIVE. However, this is a mostly benign race. Remove the racy code path. Use n_reqs to determine if a packet queue is active or not. Cc: <stable(a)vger.kernel.org> # 4.9.0 Reviewed-by: Mitko Haralanov <mitko.haralanov(a)intel.com> Reviewed-by: Mike Marciniszyn <mike.marciniszyn(a)intel.com> Signed-off-by: Michael J. Ruhl <michael.j.ruhl(a)intel.com> --- drivers/infiniband/hw/hfi1/user_sdma.c | 28 ++++++++++------------------ drivers/infiniband/hw/hfi1/user_sdma.h | 7 ++++++- 2 files changed, 16 insertions(+), 19 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/user_sdma.c b/drivers/infiniband/hw/hfi1/user_sdma.c index 619475c..4c11116 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.c +++ b/drivers/infiniband/hw/hfi1/user_sdma.c @@ -151,10 +151,6 @@ #define SDMA_REQ_HAVE_AHG 1 #define SDMA_REQ_HAS_ERROR 2 -#define SDMA_PKT_Q_INACTIVE BIT(0) -#define SDMA_PKT_Q_ACTIVE BIT(1) -#define SDMA_PKT_Q_DEFERRED BIT(2) - /* * Maximum retry attempts to submit a TX request * before putting the process to sleep. @@ -408,7 +404,6 @@ int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, struct file *fp) pq->ctxt = uctxt->ctxt; pq->subctxt = fd->subctxt; pq->n_max_reqs = hfi1_sdma_comp_ring_size; - pq->state = SDMA_PKT_Q_INACTIVE; atomic_set(&pq->n_reqs, 0); init_waitqueue_head(&pq->wait); atomic_set(&pq->n_locked, 0); @@ -491,7 +486,7 @@ int hfi1_user_sdma_free_queues(struct hfi1_filedata *fd) /* Wait until all requests have been freed. */ wait_event_interruptible( pq->wait, - (ACCESS_ONCE(pq->state) == SDMA_PKT_Q_INACTIVE)); + !atomic_read(&pq->n_reqs)); kfree(pq->reqs); kfree(pq->req_in_use); kmem_cache_destroy(pq->txreq_cache); @@ -527,6 +522,13 @@ static u8 dlid_to_selector(u16 dlid) return mapping[hash]; } +/** + * hfi1_user_sdma_process_request() - Process and start a user sdma request + * @fp: valid file pointer + * @iovec: array of io vectors to process + * @dim: overall iovec array size + * @count: number of io vector array entries processed + */ int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, unsigned long dim, unsigned long *count) { @@ -768,21 +770,13 @@ int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, } set_comp_state(pq, cq, info.comp_idx, QUEUED, 0); + pq->state = SDMA_PKT_Q_ACTIVE; /* Send the first N packets in the request to buy us some time */ ret = user_sdma_send_pkts(req, pcount); if (unlikely(ret < 0 && ret != -EBUSY)) goto free_req; /* - * It is possible that the SDMA engine would have processed all the - * submitted packets by the time we get here. Therefore, only set - * packet queue state to ACTIVE if there are still uncompleted - * requests. - */ - if (atomic_read(&pq->n_reqs)) - xchg(&pq->state, SDMA_PKT_Q_ACTIVE); - - /* * This is a somewhat blocking send implementation. * The driver will block the caller until all packets of the * request have been submitted to the SDMA engine. However, it @@ -1526,10 +1520,8 @@ static void user_sdma_txreq_cb(struct sdma_txreq *txreq, int status) static inline void pq_update(struct hfi1_user_sdma_pkt_q *pq) { - if (atomic_dec_and_test(&pq->n_reqs)) { - xchg(&pq->state, SDMA_PKT_Q_INACTIVE); + if (atomic_dec_and_test(&pq->n_reqs)) wake_up(&pq->wait); - } } static void user_sdma_free_request(struct user_sdma_request *req, bool unpin) diff --git a/drivers/infiniband/hw/hfi1/user_sdma.h b/drivers/infiniband/hw/hfi1/user_sdma.h index 3900171..09dd843 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.h +++ b/drivers/infiniband/hw/hfi1/user_sdma.h @@ -53,6 +53,11 @@ extern uint extended_psn; +enum pkt_q_sdma_state { + SDMA_PKT_Q_ACTIVE, + SDMA_PKT_Q_DEFERRED, +}; + struct hfi1_user_sdma_pkt_q { struct list_head list; unsigned ctxt; @@ -65,7 +70,7 @@ struct hfi1_user_sdma_pkt_q { struct user_sdma_request *reqs; unsigned long *req_in_use; struct iowait busy; - unsigned state; + enum pkt_q_sdma_state state; wait_queue_head_t wait; unsigned long unpinned; struct mmu_rb_handler *handler;

6 years, 9 months

1
0
0 0

[PATCH] IB/hfi1: Remove race conditions in user_sdma send path

by Michael J. Ruhl

From: Michael J. Ruhl <michael.j.ruhl(a)intel.com> commit 28a9a9e83ceae2cee25b9af9ad20d53aaa9ab951 upstream Packet queue state is over used to determine SDMA descriptor availablitity and packet queue request state. cpu 0 ret = user_sdma_send_pkts(req, pcount); cpu 0 if (atomic_read(&pq->n_reqs)) cpu 1 IRQ user_sdma_txreq_cb calls pq_update() (state to _INACTIVE) cpu 0 xchg(&pq->state, SDMA_PKT_Q_ACTIVE); At this point pq->n_reqs == 0 and pq->state is incorrectly SDMA_PKT_Q_ACTIVE. The close path will hang waiting for the state to return to _INACTIVE. This can also change the state from _DEFERRED to _ACTIVE. However, this is a mostly benign race. Remove the racy code path. Use n_reqs to determine if a packet queue is active or not. Cc: <stable(a)vger.kernel.org> # 4.14.0> Reviewed-by: Mitko Haralanov <mitko.haralanov(a)intel.com> Reviewed-by: Mike Marciniszyn <mike.marciniszyn(a)intel.com> Signed-off-by: Michael J. Ruhl <michael.j.ruhl(a)intel.com> --- drivers/infiniband/hw/hfi1/user_sdma.c | 24 ++++++++++-------------- drivers/infiniband/hw/hfi1/user_sdma.h | 9 +++++---- 2 files changed, 15 insertions(+), 18 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/user_sdma.c b/drivers/infiniband/hw/hfi1/user_sdma.c index c14ec04..cbe5ab2 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.c +++ b/drivers/infiniband/hw/hfi1/user_sdma.c @@ -187,7 +187,6 @@ int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, pq->ctxt = uctxt->ctxt; pq->subctxt = fd->subctxt; pq->n_max_reqs = hfi1_sdma_comp_ring_size; - pq->state = SDMA_PKT_Q_INACTIVE; atomic_set(&pq->n_reqs, 0); init_waitqueue_head(&pq->wait); atomic_set(&pq->n_locked, 0); @@ -276,7 +275,7 @@ int hfi1_user_sdma_free_queues(struct hfi1_filedata *fd, /* Wait until all requests have been freed. */ wait_event_interruptible( pq->wait, - (ACCESS_ONCE(pq->state) == SDMA_PKT_Q_INACTIVE)); + !atomic_read(&pq->n_reqs)); kfree(pq->reqs); kfree(pq->req_in_use); kmem_cache_destroy(pq->txreq_cache); @@ -312,6 +311,13 @@ static u8 dlid_to_selector(u16 dlid) return mapping[hash]; } +/** + * hfi1_user_sdma_process_request() - Process and start a user sdma request + * @fd: valid file descriptor + * @iovec: array of io vectors to process + * @dim: overall iovec array size + * @count: number of io vector array entries processed + */ int hfi1_user_sdma_process_request(struct hfi1_filedata *fd, struct iovec *iovec, unsigned long dim, unsigned long *count) @@ -560,21 +566,13 @@ int hfi1_user_sdma_process_request(struct hfi1_filedata *fd, req->ahg_idx = sdma_ahg_alloc(req->sde); set_comp_state(pq, cq, info.comp_idx, QUEUED, 0); + pq->state = SDMA_PKT_Q_ACTIVE; /* Send the first N packets in the request to buy us some time */ ret = user_sdma_send_pkts(req, pcount); if (unlikely(ret < 0 && ret != -EBUSY)) goto free_req; /* - * It is possible that the SDMA engine would have processed all the - * submitted packets by the time we get here. Therefore, only set - * packet queue state to ACTIVE if there are still uncompleted - * requests. - */ - if (atomic_read(&pq->n_reqs)) - xchg(&pq->state, SDMA_PKT_Q_ACTIVE); - - /* * This is a somewhat blocking send implementation. * The driver will block the caller until all packets of the * request have been submitted to the SDMA engine. However, it @@ -1391,10 +1389,8 @@ static void user_sdma_txreq_cb(struct sdma_txreq *txreq, int status) static inline void pq_update(struct hfi1_user_sdma_pkt_q *pq) { - if (atomic_dec_and_test(&pq->n_reqs)) { - xchg(&pq->state, SDMA_PKT_Q_INACTIVE); + if (atomic_dec_and_test(&pq->n_reqs)) wake_up(&pq->wait); - } } static void user_sdma_free_request(struct user_sdma_request *req, bool unpin) diff --git a/drivers/infiniband/hw/hfi1/user_sdma.h b/drivers/infiniband/hw/hfi1/user_sdma.h index 5af5233..2b5326d 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.h +++ b/drivers/infiniband/hw/hfi1/user_sdma.h @@ -94,9 +94,10 @@ #define TXREQ_FLAGS_REQ_ACK BIT(0) /* Set the ACK bit in the header */ #define TXREQ_FLAGS_REQ_DISABLE_SH BIT(1) /* Disable header suppression */ -#define SDMA_PKT_Q_INACTIVE BIT(0) -#define SDMA_PKT_Q_ACTIVE BIT(1) -#define SDMA_PKT_Q_DEFERRED BIT(2) +enum pkt_q_sdma_state { + SDMA_PKT_Q_ACTIVE, + SDMA_PKT_Q_DEFERRED, +}; /* * Maximum retry attempts to submit a TX request @@ -124,7 +125,7 @@ struct hfi1_user_sdma_pkt_q { struct user_sdma_request *reqs; unsigned long *req_in_use; struct iowait busy; - unsigned state; + enum pkt_q_sdma_state state; wait_queue_head_t wait; unsigned long unpinned; struct mmu_rb_handler *handler;

6 years, 9 months

1
0
0 0

request for 4.14-stable: c470bdc1aaf3 ("mac80211: don't WARN on bad WMM parameters from buggy APs")

by Sudip Mukherjee

Hi Greg, This was not marked for stable but seems it should be in stable. And the second patch fixes the first one. Please apply to your queue of 4.14-stable. -- Regards Sudip

6 years, 9 months

2
1
0 0

request for 4.14-stable: 2527d99789e2 ("elevator: lookup mq vs non-mq elevators")

by Sudip Mukherjee

Hi Greg, This was not marked for stable but seems it should be in stable. Please apply to your queue of 4.14-stable. -- Regards Sudip

6 years, 9 months

3
2
0 0

request for 4.14-stable: 0b8d9073539e ("netfilter: ipset: Fix wraparound in hash:*net* types")

by Sudip Mukherjee

Hi Greg, This was not marked for stable but seems it should be in stable. Please apply to your queue of 4.14-stable. -- Regards Sudip

6 years, 9 months

2
1
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror