- 03 Mar, 2016 40 commits
-
-
Joe Thornber authored
commit ed8b45a3 upstream. If dm_btree_del()'s call to push_frame() fails, e.g. due to btree_node_validator finding invalid metadata, the dm_btree_del() error path must unlock all frames (which have active dm-bufio buffers) that were pushed onto the del_stack. Otherwise, dm_bufio_client_destroy() will BUG_ON() because dm-bufio buffers have leaked, e.g.: device-mapper: bufio: leaked buffer 3, hold count 1, list 0 Signed-off-by:
Joe Thornber <ejt@redhat.com> Signed-off-by:
Mike Snitzer <snitzer@redhat.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Joe Thornber authored
commit 50dd842a upstream. When applying block operations (BOPs) do not remove them from the uncommitted BOP ring-buffer until after they've been applied -- in case we recurse. Also, perform BOP_INC operation, in dm_sm_metadata_create() and sm_metadata_extend(), in terms of the uncommitted BOP ring-buffer rather than using direct calls to sm_ll_inc(). Signed-off-by:
-
Mikulas Patocka authored
commit d98f1cd0 upstream. When I connect an Intel SSD to SATA SIL controller (PCI ID 1095:3114), any TRIM command results in I/O errors being reported in the log. There is other similar error reported with TRIM and the SIL controller: https://bugs.centos.org/view.php?id=5880 Apparently the controller doesn't support TRIM commands. This patch disables TRIM support on the SATA SIL controller. ata7.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0 ata7.00: BMDMA2 stat 0x50001 ata7.00: failed command: DATA SET MANAGEMENT ata7.00: cmd 06/01:01:00:00:00/00:00:00:00:00/a0 tag 0 dma 512 out res 51/04:01:00:00:00/00:00:00:00:00/a0 Emask 0x1 (device error) ata7.00: status: { DRDY ERR } ata7.00: error: { ABRT } ata7.00: device reported invalid CHS sector 0 sd 8:0:0:0: [sdb] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE sd 8:0:0:0: [sdb] tag#0 Sense Key : Illegal Request [current] [descriptor] sd 8:0:0:0: [sdb] tag#0 Add. Sense: Unaligned write command sd 8:0:0:0: [sdb] tag#0 CDB: Write same(16) 93 08 00 00 00 00 00 21 95 88 00 20 00 00 00 00 blk_update_request: I/O error, dev sdb, sector 2200968 Signed-off-by:
Mikulas Patocka <mpatocka@redhat.com> Signed-off-by:
Tejun Heo <tj@kernel.org> Signed-off-by:
-
Sasha Levin authored
commit 119d6f6a upstream. Because wakeups can (fundamentally) be late, a task might not be in the expected state. Therefore testing against a task's state is racy, and can yield false positives. Signed-off-by:
Sasha Levin <sasha.levin@oracle.com> Signed-off-by:
Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by:
Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: oleg@redhat.com Fixes: 9067ac85 ("wake_up_process() should be never used to wakeup a TASK_STOPPED/TRACED task") Link: http://lkml.kernel.org/r/1448933660-23082-1-git-send-email-sasha.levin@oracle.comSigned-off-by:
Ingo Molnar <mingo@kernel.org> Signed-off-by:
-
Xunlei Pang authored
commit 8295c699 upstream. root_domain::rto_mask allocated through alloc_cpumask_var() contains garbage data, this may cause problems. For instance, When doing pull_rt_task(), it may do useless iterations if rto_mask retains some extra garbage bits. Worse still, this violates the isolated domain rule for clustered scheduling using cpuset, because the tasks(with all the cpus allowed) belongs to one root domain can be pulled away into another root domain. The patch cleans the garbage by using zalloc_cpumask_var() instead of alloc_cpumask_var() for root_domain::rto_mask allocation, thereby addressing the issues. Do the same thing for root_domain's other cpumask memembers: dlo_mask, span, and online. Signed-off-by:
Xunlei Pang <xlpang@redhat.com> Signed-off-by:
-
Mirza Krak authored
commit 7cecd9ab upstream. According to SJA1000 data sheet error-warning (EI) interrupt is not cleared by setting the controller in to reset-mode. Then if we have the following case: - system is suspended (echo mem > /sys/power/state) and SJA1000 is left in operating state - A bus error condition occurs which activates EI interrupt, system is still suspended which means EI interrupt will be not be handled nor cleared. If the above two events occur, on resume there is no way to return the SJA1000 to operating state, except to cycle power to it. By simply reading the IR register on start we will clear any previous conditions that could be present. Signed-off-by:
Mirza Krak <mirza.krak@hostmobility.com> Reported-by:
Christian Magnusson <Christian.Magnusson@semcon.com> Signed-off-by:
Marc Kleine-Budde <mkl@pengutronix.de> Signed-off-by:
-
Quentin Casasnovas authored
commit 8c7188b2 upstream. Sasha's found a NULL pointer dereference in the RDS connection code when sending a message to an apparently unbound socket. The problem is caused by the code checking if the socket is bound in rds_sendmsg(), which checks the rs_bound_addr field without taking a lock on the socket. This opens a race where rs_bound_addr is temporarily set but where the transport is not in rds_bind(), leading to a NULL pointer dereference when trying to dereference 'trans' in __rds_conn_create(). Vegard wrote a reproducer for this issue, so kindly ask him to share if you're interested. I cannot reproduce the NULL pointer dereference using Vegard's reproducer with this patch, whereas I could without. Complete earlier incomplete fix to CVE-2015-6937: 74e98eb0 ("RDS: verify the underlying transport exists before creating a connection") Reviewed-by:
Vegard Nossum <vegard.nossum@oracle.com> Reviewed-by:
Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by:
Quentin Casasnovas <quentin.casasnovas@oracle.com> Signed-off-by:
David S. Miller <davem@davemloft.net> Signed-off-by:
-
Johannes Berg authored
commit c2e703a5 upstream. When using call_rcu(), the called function may be delayed quite significantly, and without a matching rcu_barrier() there's no way to be sure it has finished. Therefore, global state that could be gone/freed/reused should never be touched in the callback. Fix this in mesh by moving the atomic_dec() into the caller; that's not really a problem since we already unlinked the path and it will be destroyed anyway. This fixes a crash Jouni observed when running certain tests in a certain order, in which the mesh interface was torn down, the memory reused for a function pointer (work struct) and running that then crashed since the pointer had been decremented by 1, resulting in an invalid instruction byte stream. Fixes: eb2b9311 ("mac80211: mesh path table implementation") Reported-by:
Jouni Malinen <j@w1.fi> Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
-
Suman Anna authored
commit c13f99b7 upstream. The virtio core uses a static ida named virtio_index_ida for assigning index numbers to virtio devices during registration. The ida core may allocate some internal idr cache layers and an ida bitmap upon any ida allocation, and all these layers are truely freed only upon the ida destruction. The virtio_index_ida is not destroyed at present, leading to a memory leak when using the virtio core as a module and atleast one virtio device is registered and unregistered. Fix this by invoking ida_destroy() in the virtio core module exit. Signed-off-by:
Suman Anna <s-anna@ti.com> Signed-off-by:
Michael S. Tsirkin <mst@redhat.com> Signed-off-by:
-
Steven Rostedt (Red Hat) authored
commit b81f472a upstream. Do not update the read stamp after swapping out the reader page from the write buffer. If the reader page is swapped out of the buffer before an event is written to it, then the read_stamp may get an out of date timestamp, as the page timestamp is updated on the first commit to that page. rb_get_reader_page() only returns a page if it has an event on it, otherwise it will return NULL. At that point, check if the page being returned has events and has not been read yet. Then at that point update the read_stamp to match the time stamp of the reader page. Signed-off-by:
Steven Rostedt <rostedt@goodmis.org> Signed-off-by:
-
Jan Engelhardt authored
commit d94e5a61 upstream. target_core_sbc's compare_and_write functionality suffers from taking data at the wrong memory location when writing a CAW request to disk when a SGL offset is non-zero. This can happen with loopback and vhost-scsi fabric drivers when SCF_PASSTHROUGH_SG_TO_MEM_NOALLOC is used to map existing user-space SGL memory into COMPARE_AND_WRITE READ/WRITE payload buffers. Given the following sample LIO subtopology, % targetcli ls /loopback/ o- loopback ................................. [1 Target] o- naa.6001405ebb8df14a ....... [naa.60014059143ed2b3] o- luns ................................... [2 LUNs] o- lun0 ................ [iblock/ram0 (/dev/ram0)] o- lun1 ................ [iblock/ram1 (/dev/ram1)] % lsscsi -g [3:0:1:0] disk LIO-ORG IBLOCK 4.0 /dev/sdc /dev/sg3 [3:0:1:1] disk LIO-ORG IBLOCK 4.0 /dev/sdd /dev/sg4 the following bug can be observed in Linux 4.3 and 4.4~rc1: % perl -e 'print chr$_ for 0..255,reverse 0..255' >rand % perl -e 'print "\0" x 512' >zero % cat rand >/dev/sdd % sg_compare_and_write -i rand -D zero --lba 0 /dev/sdd % sg_compare_and_write -i zero -D rand --lba 0 /dev/sdd Miscompare reported % hexdump -Cn 512 /dev/sdd 00000000 0f 0e 0d 0c 0b 0a 09 08 07 06 05 04 03 02 01 00 00000010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 * 00000200 Rather than writing all-zeroes as instructed with the -D file, it corrupts the data in the sector by splicing some of the original bytes in. The page of the first entry of cmd->t_data_sg includes the CDB, and sg->offset is set to a position past the CDB. I presume that sg->offset is also the right choice to use for subsequent sglist members. Signed-off-by:
Jan Engelhardt <jengelh@netitwork.de> Tested-by:
Douglas Gilbert <dgilbert@interlog.com> Signed-off-by:
Nicholas Bellinger <nab@linux-iscsi.org> Signed-off-by:
-
Nicholas Bellinger authored
commit 057085e5 upstream. This patch addresses a race + use after free where the first stage of COMPARE_AND_WRITE in compare_and_write_callback() is rescheduled after the backend sends the secondary WRITE, resulting in second stage compare_and_write_post() callback completing in target_complete_ok_work() before the first can return. Because current code depends on checking se_cmd->se_cmd_flags after return from se_cmd->transport_complete_callback(), this results in first stage having SCF_COMPARE_AND_WRITE_POST set, which incorrectly falls through into second stage CAW processing code, eventually triggering a NULL pointer dereference due to use after free. To address this bug, pass in a new *post_ret parameter into se_cmd->transport_complete_callback(), and depend upon this value instead of ->se_cmd_flags to determine when to return or fall through into ->queue_status() code for CAW. Cc: Sagi Grimberg <sagig@mellanox.com> Signed-off-by:
-
Jan Kara authored
commit c2489e07 upstream. The following test program from Dmitry can cause softlockups or RCU stalls as it copies 1GB from tmpfs into eventfd and we don't have any scheduling point at that path in sendfile(2) implementation: int r1 = eventfd(0, 0); int r2 = memfd_create("", 0); unsigned long n = 1<<30; fallocate(r2, 0, 0, n); sendfile(r1, r2, 0, n); Add cond_resched() into __splice_from_pipe() to fix the problem. CC: Dmitry Vyukov <dvyukov@google.com> Signed-off-by:
Jan Kara <jack@suse.cz> Signed-off-by:
Al Viro <viro@zeniv.linux.org.uk> Signed-off-by:
-
Vineet Gupta authored
commit 2e22502c upstream. Fixes STAR 9000953410: "perf callgraph profiling causing RCU stalls" | perf record -g -c 15000 -e cycles /sbin/hackbench | | INFO: rcu_preempt self-detected stall on CPU | 1: (1 GPs behind) idle=609/140000000000002/0 softirq=2914/2915 fqs=603 | Task dump for CPU 1: in-kernel dwarf unwinder has a fast binary lookup and a fallback linear search (which iterates thru each of ~11K entries) thus takes 2 orders of magnitude longer (~3 million cycles vs. 2000). Routines written in hand assembler lack dwarf info (as we don't support assembler CFI pseudo-ops yet) fail the unwinder binary lookup, hit linear search, failing nevertheless in the end. However the linear search is pointless as binary lookup tables are created from it in first place. It is impossible to have binary lookup fail while succeed the linear search. It is pure waste of cycles thus removed by this patch. This manifested as RCU stalls / NMI watchdog splat when running hackbench under perf with callgraph profiling. The triggering condition was perf counter overflowing in routine lacking dwarf info (like memset) leading to patheic 3 million cycle unwinder slow path and by the time it returned new interrupts were already pending (Timer, IPI) and taken rightaway. The original memset didn't make forward progress, system kept accruing more interrupts and more unwinder delayes in a vicious feedback loop, ultimately triggering the NMI diagnostic. Signed-off-by:
Vineet Gupta <vgupta@synopsys.com> Signed-off-by:
-
Kees Cook authored
commit 02e2a5bf upstream. If md->signature == MAC_DRIVER_MAGIC and md->block_size == 1023, a single 512 byte sector would be read (secsize / 512). However the partition structure would be located past the end of the buffer (secsize % 512). Signed-off-by:
Kees Cook <keescook@chromium.org> Signed-off-by:
Jens Axboe <axboe@fb.com> Signed-off-by:
-
Luca Porzio authored
commit d3df0465 upstream. Anytime a write operation is performed with Reliable Write flag enabled, the eMMC device is enforced to bypass the cache and do a write to the underling NVM device by Jedec specification; this causes a performance penalty since write operations can't be optimized by the device cache. In our tests, we replayed a typical mobile daily trace pattern and found ~9% overall time reduction in trace replay by using this patch. Also the write ops within 4KB~64KB chunk size range get a 40~60% performance improvement by using the patch (as this range of write chunks are the ones affected by REQ_META). This patch has been discussed in the Mobile & Embedded Linux Storage Forum and it's the results of feedbacks from many people. We also checked with fsdevl and f2fs mailing list developers that this change in the usage of REQ_META is not affecting FS behavior and we got positive feedbacks. Reporting here the feedbacks: http://comments.gmane.org/gmane.linux.file-systems/97219 http://thread.gmane.org/gmane.linux.file-systems.f2fs/3178/focus=3183Signed-off-by:
Bruce Ford <bford@micron.com> Signed-off-by:
Luca Porzio <lporzio@micron.com> Fixes: ce39f9d1 ("mmc: support packed write command for eMMC4.5 devices") Signed-off-by:
Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by:
-
K. Y. Srinivasan authored
commit 8cf308e1 upstream. Don't set the SRB_FLAGS_QUEUE_ACTION_ENABLE flag since we are not specifying tags. Without this, the qlogic driver doesn't work properly with storvsc. Signed-off-by:
K. Y. Srinivasan <kys@microsoft.com> Signed-off-by:
James Bottomley <JBottomley@Odin.com> Signed-off-by:
-
sumit.saxena@avagotech.com authored
commit 323c4a02 upstream. This is an issue on SMAP enabled CPUs and 32 bit apps running on 64 bit OS. Do not access user memory from kernel code. The SMAP bit restricts accessing user memory from kernel code. Signed-off-by:
Sumit Saxena <sumit.saxena@avagotech.com> Signed-off-by:
Kashyap Desai <kashyap.desai@avagotech.com> Reviewed-by:
Tomas Henzl <thenzl@redhat.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by:
-
sumit.saxena@avagotech.com authored
commit 357ae967 upstream. Do not use PAGE_SIZE marco to calculate max_sectors per I/O request. Driver code assumes PAGE_SIZE will be always 4096 which can lead to wrongly calculated value if PAGE_SIZE is not 4096. This issue was reported in Ubuntu Bugzilla Bug #1475166. Signed-off-by:
-
Andy Shevchenko authored
commit 39416677 upstream. We replace __fls() by __ffs() since we have to find a *minimum* data width that satisfies both source and destination. While here, rename dwc_fast_fls() to dwc_fast_ffs() which it really is. Fixes: 4c2d56c5 (dw_dmac: introduce dwc_fast_fls()) Signed-off-by:
Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by:
Vinod Koul <vinod.koul@intel.com> Signed-off-by:
-
Valentin Rothberg authored
commit 90adf98d upstream. Since commit 1c6c6952 ("genirq: Reject bogus threaded irq requests") threaded IRQs without a primary handler need to be requested with IRQF_ONESHOT, otherwise the request will fail. scripts/coccinelle/misc/irqf_oneshot.cocci detected this issue. Fixes: b5874f33 ("wm831x_power: Use genirq") Signed-off-by:
Valentin Rothberg <valentinrothberg@gmail.com> Signed-off-by:
Sebastian Reichel <sre@kernel.org> Signed-off-by:
-
Dan Carpenter authored
commit 1f35d04a upstream. The iomap[] array has PCIM_IOMAP_MAX (6) elements and not DEVICE_COUNT_RESOURCE (16). This bug was found using a static checker. It may be that the "if (!(mask & (1 << i)))" check means we never actually go past the end of the array in real life. Fixes: ec04b075 ('iomap: implement pcim_iounmap_regions()') Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Acked-by:
-
Andrey Ryabinin authored
commit 0ad95472 upstream. Commit cb7323ff ("lockd: create and use per-net NSM RPC clients on MON/UNMON requests") introduced per-net NSM RPC clients. Unfortunately this doesn't make any sense without per-net nsm_handle. E.g. the following scenario could happen Two hosts (X and Y) in different namespaces (A and B) share the same nsm struct. 1. nsm_monitor(host_X) called => NSM rpc client created, nsm->sm_monitored bit set. 2. nsm_mointor(host-Y) called => nsm->sm_monitored already set, we just exit. Thus in namespace B ln->nsm_clnt == NULL. 3. host X destroyed => nsm->sm_count decremented to 1 4. host Y destroyed => nsm_unmonitor() => nsm_mon_unmon() => NULL-ptr dereference of *ln->nsm_clnt So this could be fixed by making per-net nsm_handles list, instead of global. Thus different net namespaces will not be able share the same nsm_handle. Signed-off-by:
Andrey Ryabinin <aryabinin@virtuozzo.com> Signed-off-by:
J. Bruce Fields <bfields@redhat.com> Signed-off-by:
-
Alex Deucher authored
commit 4e7697ed upstream. On some cards it takes a relatively long time for the change to take place. Make a timeout non-fatal. bug: https://bugs.freedesktop.org/show_bug.cgi?id=76130Signed-off-by:
Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
-
Alex Deucher authored
commit 24dd2f64 upstream. Avoids spew on resume for systems where sysfs may fail even on init. bug: https://bugzilla.kernel.org/show_bug.cgi?id=106851Signed-off-by:
-
NeilBrown authored
commit b02bab6b upstream. These async_XX functions are called from md/raid5 in an atomic section, between get_cpu() and put_cpu(), so they must not sleep. So use GFP_NOWAIT rather than GFP_IO. Dan Williams writes: Longer term async_tx needs to be merged into md directly as we can allocate this unmap data statically per-stripe rather than per request. Fixed: 7476bd79 ("async_pq: convert to dmaengine_unmap_data") Reported-and-tested-by:
Stanislav Samsonov <slava@annapurnalabs.com> Acked-by:
Dan Williams <dan.j.williams@intel.com> Signed-off-by:
NeilBrown <neilb@suse.com> Signed-off-by:
-
Roman Volkov authored
commit f9eccf24 upstream. The vt8500 clocksource driver declares itself as capable to handle the minimum delay of 4 cycles by passing the value into clockevents_config_and_register(). The vt8500_timer_set_next_event() requires the passed cycles value to be at least 16. The impact is that userspace hangs in nanosleep() calls with small delay intervals. This problem is reproducible in Linux 4.2 starting from: c6eb3f70 ('hrtimer: Get rid of hrtimer softirq') From Russell King, more detailed explanation: "It's a speciality of the StrongARM/PXA hardware. It takes a certain number of OSCR cycles for the value written to hit the compare registers. So, if a very small delta is written (eg, the compare register is written with a value of OSCR + 1), the OSCR will have incremented past this value before it hits the underlying hardware. The result is, that you end up waiting a very long time for the OSCR to wrap before the event fires. So, we introduce a check in set_next_event() to detect this and return -ETIME if the calculated delta is too small, which causes the generic clockevents code to retry after adding the min_delta specified in clockevents_config_and_register() to the current time value. min_delta must be sufficient that we don't re-trip the -ETIME check - if we do, we will return -ETIME, forward the next event time, try to set it, return -ETIME again, and basically lock the system up. So, min_delta must be larger than the check inside set_next_event(). A factor of two was chosen to ensure that this situation would never occur. The PXA code worked on PXA systems for years, and I'd suggest no one changes this mechanism without access to a wide range of PXA systems, otherwise they're risking breakage." Cc: Russell King <linux@arm.linux.org.uk> Acked-by:
Alexey Charkov <alchark@gmail.com> Signed-off-by:
Roman Volkov <rvolkov@v1ros.org> Signed-off-by:
Daniel Lezcano <daniel.lezcano@linaro.org> Signed-off-by:
-
Roman Volkov authored
commit 0f090bf1 upstream. Since WM8650 has the same 'WMT' SDHC controller as WM8505, and the driver is already in the kernel, this node enables the controller support for WM8650 Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
-
Thomas Gleixner authored
commit abc7e40c upstream. If a interrupt chip utilizes chip->buslock then free_irq() can deadlock in the following way: CPU0 CPU1 interrupt(X) (Shared or spurious) free_irq(X) interrupt_thread(X) chip_bus_lock(X) irq_finalize_oneshot(X) chip_bus_lock(X) synchronize_irq(X) synchronize_irq() waits for the interrupt thread to complete, i.e. forever. Solution is simple: Drop chip_bus_lock() before calling synchronize_irq() as we do with the irq_desc lock. There is nothing to be protected after the point where irq_desc lock has been released. This adds chip_bus_lock/unlock() to the remove_irq() code path, but that's actually correct in the case where remove_irq() is called on such an interrupt. The current users of remove_irq() are not affected as none of those interrupts is on a chip which requires buslock. Reported-by:
Fredrik Markström <fredrik.markstrom@gmail.com> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
-
Peter Zijlstra authored
commit 9916e214 upstream. Remove the direct {push,pull} balancing operations from switched_{from,to}_dl() / prio_changed_dl() and use the balance callback queue. Again, err on the side of too many reschedules; since too few is a hard bug while too many is just annoying. Signed-off-by:
Byungchul Park <byungchul.park@lge.com> Signed-off-by:
-
Peter Zijlstra authored
commit 0ea60c20 upstream. In order to be able to use pull_dl_task() from a callback, we need to do away with the return value. Since the return value indicates if we should reschedule, do this inside the function. Since not all callers currently do this, this can increase the number of reschedules due rt balancing. Too many reschedules is not a correctness issues, too few are. Signed-off-by:
-
Peter Zijlstra authored
commit fd7a4bed upstream. Remove the direct {push,pull} balancing operations from switched_{from,to}_rt() / prio_changed_rt() and use the balance callback queue. Again, err on the side of too many reschedules; since too few is a hard bug while too many is just annoying. Signed-off-by:
-
Peter Zijlstra authored
commit 8046d680 upstream. In order to be able to use pull_rt_task() from a callback, we need to do away with the return value. Since the return value indicates if we should reschedule, do this inside the function. Since not all callers currently do this, this can increase the number of reschedules due rt balancing. Too many reschedules is not a correctness issues, too few are. Signed-off-by:
-
Peter Zijlstra authored
commit 4c9a4bc8 upstream. In order to remove dropping rq->lock from the switched_{to,from}()/prio_changed() sched_class methods, run the balance callbacks after it. We need to remove dropping rq->lock because its buggy, suppose using sched_setattr()/sched_setscheduler() to change a running task from FIFO to OTHER. By the time we get to switched_from_rt() the task is already enqueued on the cfs runqueues. If switched_from_rt() does pull_rt_task() and drops rq->lock, load-balancing can come in and move our task @p to another rq. The subsequent switched_to_fair() still assumes @p is on @rq and bad things will happen. By using balance callbacks we delay the load-balancing operations {rt,dl}x{push,pull} until we've done all the important work and the task is fully set up. Furthermore, the balance callbacks do not know about @p, therefore they cannot get confused like this. Reported-by:
Mike Galbraith <umgwanakikbuti@gmail.com> Signed-off-by:
-
Peter Zijlstra authored
commit e3fca9e7 upstream. Generalize the post_schedule() stuff into a balance callback list. This allows us to more easily use it outside of schedule() and cross sched_class. Signed-off-by:
-
Peter Zijlstra authored
commit 6c3b4d44 upstream. The idle post_schedule flag is just a vile waste of time, furthermore it appears unneeded, move the idle_enter_fair() call into pick_next_task_idle(). Signed-off-by:
-
Hannes Frederic Sowa authored
commit 415e3d3e upstream. The commit referenced in the Fixes tag incorrectly accounted the number of in-flight fds over a unix domain socket to the original opener of the file-descriptor. This allows another process to arbitrary deplete the original file-openers resource limit for the maximum of open files. Instead the sending processes and its struct cred should be credited. To do so, we add a reference counted struct user_struct pointer to the scm_fp_list and use it to account for the number of inflight unix fds. Fixes: 712f4aad ("unix: properly account for FDs passed over unix sockets") Reported-by:
David Herrmann <dh.herrmann@gmail.com> Cc: David Herrmann <dh.herrmann@gmail.com> Cc: Willy Tarreau <w@1wt.eu> Cc: Linus Torvalds <torvalds@linux-foundation.org> Suggested-by:
Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by:
-
Manish Chopra authored
commit ad6afbe9 upstream. The scratchpad is a shared block between all functions of a given device. Due to HW limitations, we can't properly close its parity notifications to all functions on legal flows. E.g., it's possible that while taking a register dump from one function a parity error would be triggered on other functions. Today driver doesn't consider this parity as a 'real' parity unless its being accompanied by additional indications [which would happen in a real parity scenario]; But it does print notifications for such events in the system logs. This eliminates such prints - in case of real parities driver would have additional indications; But if this is the only signal user will not even see a parity being logged in the system. Signed-off-by:
Manish Chopra <Manish.Chopra@qlogic.com> Signed-off-by:
Yuval Mintz <Yuval.Mintz@qlogic.com> Signed-off-by:
Ariel Elior <Ariel.Elior@qlogic.com> Signed-off-by:
Patrick Schaaf <netdev@bof.de> Tested-by:
-
Olga Kornievskaia authored
commit a41cbe86 upstream. A test case is as the description says: open(foobar, O_WRONLY); sleep() --> reboot the server close(foobar) The bug is because in nfs4state.c in nfs4_reclaim_open_state() a few line before going to restart, there is clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &state->flags). NFS4CLNT_RECLAIM_NOGRACE is a flag for the client states not open owner states. Value of NFS4CLNT_RECLAIM_NOGRACE is 4 which is the value of NFS_O_WRONLY_STATE in nfs4_state->flags. So clearing it wipes out state and when we go to close it, “call_close” doesn’t get set as state flag is not set and CLOSE doesn’t go on the wire. Signed-off-by:
Olga Kornievskaia <aglo@umich.edu> Signed-off-by:
Trond Myklebust <trond.myklebust@primarydata.com> Signed-off-by:
-
Christophe Leroy authored
commit 0ff28d9f upstream. Using sendfile with below small program to get MD5 sums of some files, it appear that big files (over 64kbytes with 4k pages system) get a wrong MD5 sum while small files get the correct sum. This program uses sendfile() to send a file to an AF_ALG socket for hashing. /* md5sum2.c */ #include <stdio.h> #include <stdlib.h> #include <unistd.h> #include <string.h> #include <fcntl.h> #include <sys/socket.h> #include <sys/stat.h> #include <sys/types.h> #include <linux/if_alg.h> int main(int argc, char **argv) { int sk = socket(AF_ALG, SOCK_SEQPACKET, 0); struct stat st; struct sockaddr_alg sa = { .salg_family = AF_ALG, .salg_type = "hash", .salg_name = "md5", }; int n; bind(sk, (struct sockaddr*)&sa, sizeof(sa)); for (n = 1; n < argc; n++) { int size; int offset = 0; char buf[4096]; int fd; int sko; int i; fd = open(argv[n], O_RDONLY); sko = accept(sk, NULL, 0); fstat(fd, &st); size = st.st_size; sendfile(sko, fd, &offset, size); size = read(sko, buf, sizeof(buf)); for (i = 0; i < size; i++) printf("%2.2x", buf[i]); printf(" %s\n", argv[n]); close(fd); close(sko); } exit(0); } Test below is done using official linux patch files. First result is with a software based md5sum. Second result is with the program above. root@vgoip:~# ls -l patch-3.6.* -rw-r--r-- 1 root root 64011 Aug 24 12:01 patch-3.6.2.gz -rw-r--r-- 1 root root 94131 Aug 24 12:01 patch-3.6.3.gz root@vgoip:~# md5sum patch-3.6.* b3ffb9848196846f31b2ff133d2d6443 patch-3.6.2.gz c5e8f687878457db77cb7158c38a7e43 patch-3.6.3.gz root@vgoip:~# ./md5sum2 patch-3.6.* b3ffb9848196846f31b2ff133d2d6443 patch-3.6.2.gz 5fd77b24e68bb24dcc72d6e57c64790e patch-3.6.3.gz After investivation, it appears that sendfile() sends the files by blocks of 64kbytes (16 times PAGE_SIZE). The problem is that at the end of each block, the SPLICE_F_MORE flag is missing, therefore the hashing operation is reset as if it was the end of the file. This patch adds SPLICE_F_MORE to the flags when more data is pending. With the patch applied, we get the correct sums: root@vgoip:~# md5sum patch-3.6.* b3ffb9848196846f31b2ff133d2d6443 patch-3.6.2.gz c5e8f687878457db77cb7158c38a7e43 patch-3.6.3.gz root@vgoip:~# ./md5sum2 patch-3.6.* b3ffb9848196846f31b2ff133d2d6443 patch-3.6.2.gz c5e8f687878457db77cb7158c38a7e43 patch-3.6.3.gz Signed-off-by:
Christophe Leroy <christophe.leroy@c-s.fr> Signed-off-by:
-
