- 04 May, 2016 40 commits
-
-
Steve Capper authored
commit 66ee95d1 upstream. HugeTLB pages cannot be split, so we use the compound_mapcount to track rmaps. Currently page_mapped() will check the compound_mapcount, but will also go through the constituent pages of a THP compound page and query the individual _mapcount's too. Unfortunately, page_mapped() does not distinguish between HugeTLB and THP compound pages and assumes that a compound page always needs to have HPAGE_PMD_NR pages querying. For most cases when dealing with HugeTLB this is just inefficient, but for scenarios where the HugeTLB page size is less than the pmd block size (e.g. when using contiguous bit on ARM) this can lead to crashes. This patch adjusts the page_mapped function such that we skip the unnecessary THP reference checks for HugeTLB pages. Fixes: e1534ae9 ("mm: differentiate page_mapped() from page_mapcount() for compound pages") Signed-off-by:
Steve Capper <steve.capper@arm.com> Acked-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Will Deacon <will.deacon@arm.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Michal Hocko <mhocko@suse.com> Cc: Ingo Molnar <mingo@kernel.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tejun Heo authored
commit 264a0ae1 upstream. Hello, So, this ended up a lot simpler than I originally expected. I tested it lightly and it seems to work fine. Petr, can you please test these two patches w/o the lru drain drop patch and see whether the problem is gone? Thanks. ------ 8< ------ If charge moving is used, memcg performs relabeling of the affected pages from its ->attach callback which is called under both cgroup_threadgroup_rwsem and thus can't create new kthreads. This is fragile as various operations may depend on workqueues making forward progress which relies on the ability to create new kthreads. There's no reason to perform charge moving from ->attach which is deep in the task migration path. Move it to ->post_attach which is called after the actual migration is finished and cgroup_threadgroup_rwsem is dropped. * move_charge_struct->mm is added and ->can_attach is now responsible for pinning and recording the target mm. mem_cgroup_clear_mc() is updated accordingly. This also simplifies mem_cgroup_move_task(). * mem_cgroup_move_task() is now called from ->post_attach instead of ->attach. Signed-off-by:
Tejun Heo <tj@kernel.org> Cc: Johannes Weiner <hannes@cmpxchg.org> Acked-by:
Michal Hocko <mhocko@kernel.org> Debugged-and-tested-by:
Petr Mladek <pmladek@suse.com> Reported-by:
Cyril Hrubis <chrubis@suse.cz> Reported-by:
Johannes Weiner <hannes@cmpxchg.org> Fixes: 1ed13287 ("sched, cgroup: replace signal_struct->group_rwsem with a global percpu_rwsem") Signed-off-by:
-
Tejun Heo authored
commit 5cf1cacb upstream. Since e93ad19d ("cpuset: make mm migration asynchronous"), cpuset kicks off asynchronous NUMA node migration if necessary during task migration and flushes it from cpuset_post_attach_flush() which is called at the end of __cgroup_procs_write(). This is to avoid performing migration with cgroup_threadgroup_rwsem write-locked which can lead to deadlock through dependency on kworker creation. memcg has a similar issue with charge moving, so let's convert it to an official callback rather than the current one-off cpuset specific function. This patch adds cgroup_subsys->post_attach callback and makes cpuset register cpuset_post_attach_flush() as its ->post_attach. The conversion is mostly one-to-one except that the new callback is called under cgroup_mutex. This is to guarantee that no other migration operations are started before ->post_attach callbacks are finished. cgroup_mutex is one of the outermost mutex in the system and has never been and shouldn't be a problem. We can add specialized synchronization around __cgroup_procs_write() but I don't think there's any noticeable benefit. Signed-off-by:
-
Jesper Dangaard Brouer authored
commit 376bf125 upstream. This change is primarily an attempt to make it easier to realize the optimizations the compiler performs in-case CONFIG_MEMCG_KMEM is not enabled. Performance wise, even when CONFIG_MEMCG_KMEM is compiled in, the overhead is zero. This is because, as long as no process have enabled kmem cgroups accounting, the assignment is replaced by asm-NOP operations. This is possible because memcg_kmem_enabled() uses a static_key_false() construct. It also helps readability as it avoid accessing the p[] array like: p[size - 1] which "expose" that the array is processed backwards inside helper function build_detached_freelist(). Lastly this also makes the code more robust, in error case like passing NULL pointers in the array. Which were previously handled before commit 03374518 ("slub: add missing kmem cgroup support to kmem_cache_free_bulk"). Fixes: 03374518 ("slub: add missing kmem cgroup support to kmem_cache_free_bulk") Signed-off-by:
Jesper Dangaard Brouer <brouer@redhat.com> Cc: Christoph Lameter <cl@linux.com> Cc: Pekka Enberg <penberg@kernel.org> Cc: David Rientjes <rientjes@google.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Vladimir Davydov <vdavydov@virtuozzo.com> Signed-off-by:
-
Roman Pen authored
commit 346c09f8 upstream. The bug in a workqueue leads to a stalled IO request in MQ ctx->rq_list with the following backtrace: [ 601.347452] INFO: task kworker/u129:5:1636 blocked for more than 120 seconds. [ 601.347574] Tainted: G O 4.4.5-1-storage+ #6 [ 601.347651] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 601.348142] kworker/u129:5 D ffff880803077988 0 1636 2 0x00000000 [ 601.348519] Workqueue: ibnbd_server_fileio_wq ibnbd_dev_file_submit_io_worker [ibnbd_server] [ 601.348999] ffff880803077988 ffff88080466b900 ffff8808033f9c80 ffff880803078000 [ 601.349662] ffff880807c95000 7fffffffffffffff ffffffff815b0920 ffff880803077ad0 [ 601.350333] ffff8808030779a0 ffffffff815b01d5 0000000000000000 ffff880803077a38 [ 601.350965] Call Trace: [ 601.351203] [<ffffffff815b0920>] ? bit_wait+0x60/0x60 [ 601.351444] [<ffffffff815b01d5>] schedule+0x35/0x80 [ 601.351709] [<ffffffff815b2dd2>] schedule_timeout+0x192/0x230 [ 601.351958] [<ffffffff812d43f7>] ? blk_flush_plug_list+0xc7/0x220 [ 601.352208] [<ffffffff810bd737>] ? ktime_get+0x37/0xa0 [ 601.352446] [<ffffffff815b0920>] ? bit_wait+0x60/0x60 [ 601.352688] [<ffffffff815af784>] io_schedule_timeout+0xa4/0x110 [ 601.352951] [<ffffffff815b3a4e>] ? _raw_spin_unlock_irqrestore+0xe/0x10 [ 601.353196] [<ffffffff815b093b>] bit_wait_io+0x1b/0x70 [ 601.353440] [<ffffffff815b056d>] __wait_on_bit+0x5d/0x90 [ 601.353689] [<ffffffff81127bd0>] wait_on_page_bit+0xc0/0xd0 [ 601.353958] [<ffffffff81096db0>] ? autoremove_wake_function+0x40/0x40 [ 601.354200] [<ffffffff81127cc4>] __filemap_fdatawait_range+0xe4/0x140 [ 601.354441] [<ffffffff81127d34>] filemap_fdatawait_range+0x14/0x30 [ 601.354688] [<ffffffff81129a9f>] filemap_write_and_wait_range+0x3f/0x70 [ 601.354932] [<ffffffff811ced3b>] blkdev_fsync+0x1b/0x50 [ 601.355193] [<ffffffff811c82d9>] vfs_fsync_range+0x49/0xa0 [ 601.355432] [<ffffffff811cf45a>] blkdev_write_iter+0xca/0x100 [ 601.355679] [<ffffffff81197b1a>] __vfs_write+0xaa/0xe0 [ 601.355925] [<ffffffff81198379>] vfs_write+0xa9/0x1a0 [ 601.356164] [<ffffffff811c59d8>] kernel_write+0x38/0x50 The underlying device is a null_blk, with default parameters: queue_mode = MQ submit_queues = 1 Verification that nullb0 has something inflight: root@pserver8:~# cat /sys/block/nullb0/inflight 0 1 root@pserver8:~# find /sys/block/nullb0/mq/0/cpu* -name rq_list -print -exec cat {} \; ... /sys/block/nullb0/mq/0/cpu2/rq_list CTX pending: ffff8838038e2400 ... During debug it became clear that stalled request is always inserted in the rq_list from the following path: save_stack_trace_tsk + 34 blk_mq_insert_requests + 231 blk_mq_flush_plug_list + 281 blk_flush_plug_list + 199 wait_on_page_bit + 192 __filemap_fdatawait_range + 228 filemap_fdatawait_range + 20 filemap_write_and_wait_range + 63 blkdev_fsync + 27 vfs_fsync_range + 73 blkdev_write_iter + 202 __vfs_write + 170 vfs_write + 169 kernel_write + 56 So blk_flush_plug_list() was called with from_schedule == true. If from_schedule is true, that means that finally blk_mq_insert_requests() offloads execution of __blk_mq_run_hw_queue() and uses kblockd workqueue, i.e. it calls kblockd_schedule_delayed_work_on(). That means, that we race with another CPU, which is about to execute __blk_mq_run_hw_queue() work. Further debugging shows the following traces from different CPUs: CPU#0 CPU#1 ---------------------------------- ------------------------------- reqeust A inserted STORE hctx->ctx_map[0] bit marked kblockd_schedule...() returns 1 <schedule to kblockd workqueue> request B inserted STORE hctx->ctx_map[1] bit marked kblockd_schedule...() returns 0 *** WORK PENDING bit is cleared *** flush_busy_ctxs() is executed, but bit 1, set by CPU#1, is not observed As a result request B pended forever. This behaviour can be explained by speculative LOAD of hctx->ctx_map on CPU#0, which is reordered with clear of PENDING bit and executed _before_ actual STORE of bit 1 on CPU#1. The proper fix is an explicit full barrier <mfence>, which guarantees that clear of PENDING bit is to be executed before all possible speculative LOADS or STORES inside actual work function. Signed-off-by:
Roman Pen <roman.penyaev@profitbricks.com> Cc: Gioh Kim <gi-oh.kim@profitbricks.com> Cc: Michael Wang <yun.wang@profitbricks.com> Cc: Tejun Heo <tj@kernel.org> Cc: Jens Axboe <axboe@kernel.dk> Cc: linux-block@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by:
-
Keith Busch authored
commit 1bdb8970 upstream. If x86_vector_alloc_irq() fails x86_vector_free_irqs() is invoked to cleanup the already allocated vectors. This subsequently calls clear_vector_irq(). The failed irq has no vector assigned, which triggers the BUG_ON(!vector) in clear_vector_irq(). We cannot suppress the call to x86_vector_free_irqs() for the failed interrupt, because the other data related to this irq must be cleaned up as well. So calling clear_vector_irq() with vector == 0 is legitimate. Remove the BUG_ON and return if vector is zero, [ tglx: Massaged changelog ] Fixes: b5dc8e6c "x86/irq: Use hierarchical irqdomain to manage CPU interrupt vectors" Signed-off-by:
Keith Busch <keith.busch@intel.com> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
-
Ard Biesheuvel authored
commit 9c6672ac upstream. Commit 6d80dba1 ("efi: Provide a non-blocking SetVariable() operation") implemented a non-blocking alternative for the UEFI SetVariable() invocation performed by efivars, since it may occur in atomic context. However, this version of the function was never exposed via the efivars struct, so the non-blocking versions was not actually callable. Fix that. Signed-off-by:
Ard Biesheuvel <ard.biesheuvel@linaro.org> Signed-off-by:
Matt Fleming <matt@codeblueprint.co.uk> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: linux-efi@vger.kernel.org Fixes: 6d80dba1 ("efi: Provide a non-blocking SetVariable() operation") Link: http://lkml.kernel.org/r/1454364428-494-2-git-send-email-matt@codeblueprint.co.ukSigned-off-by:
Ingo Molnar <mingo@kernel.org> Signed-off-by:
-
Laszlo Ersek authored
commit 630ba0cc upstream. The variable_matches() function can currently read "var_name[len]", for example when: - var_name[0] == 'a', - len == 1 - match_name points to the NUL-terminated string "ab". This function is supposed to accept "var_name" inputs that are not NUL-terminated (hence the "len" parameter"). Document the function, and access "var_name[*match]" only if "*match" is smaller than "len". Reported-by:
Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by:
Laszlo Ersek <lersek@redhat.com> Cc: Peter Jones <pjones@redhat.com> Cc: Matthew Garrett <mjg59@coreos.com> Cc: Jason Andryuk <jandryuk@gmail.com> Cc: Jani Nikula <jani.nikula@linux.intel.com> Link: http://thread.gmane.org/gmane.comp.freedesktop.xorg.drivers.intel/86906Signed-off-by:
-
Jason Gunthorpe authored
commit e6bd18f5 upstream. The drivers/infiniband stack uses write() as a replacement for bi-directional ioctl(). This is not safe. There are ways to trigger write calls that result in the return structure that is normally written to user space being shunted off to user specified kernel memory instead. For the immediate repair, detect and deny suspicious accesses to the write API. For long term, update the user space libraries and the kernel API to something that doesn't present the same security vulnerabilities (likely a structured ioctl() interface). The impacted uAPI interfaces are generally only available if hardware from drivers/infiniband is installed in the system. Reported-by:
Jann Horn <jann@thejh.net> Signed-off-by:
Jason Gunthorpe <jgunthorpe@obsidianresearch.com> [ Expanded check to all known write() entry points ] Signed-off-by:
Doug Ledford <dledford@redhat.com> Signed-off-by:
-
Sagi Grimberg authored
commit 986ef95e upstream. mlx5 devices (Connect-IB, ConnectX-4, ConnectX-4-LX) has a limitation where rdma read work queue entries cannot exceed 512 bytes. A rdma_read wqe needs to fit in 512 bytes: - wqe control segment (16 bytes) - rdma segment (16 bytes) - scatter elements (16 bytes each) So max_sge_rd should be: (512 - 16 - 16) / 16 = 30. Reported-by:
Christoph Hellwig <hch@lst.de> Tested-by:
Sagi Grimberg <sagig@grimberg.me> Signed-off-by:
Leon Romanovsky <leonro@mellanox.com> Signed-off-by:
-
chunfan chen authored
commit dc386ce7 upstream. The port_open flag is not applicable for IBSS mode. IBSS data path was broken when port_open flag was introduced. This patch fixes the problem by correcting the checks. Fixes: 5c894633 ("mwifiex: enable traffic only when port is open") Signed-off-by:
chunfan chen <jeffc@marvell.com> Signed-off-by:
Amitkumar Karwar <akarwar@marvell.com> Signed-off-by:
Kalle Valo <kvalo@codeaurora.org> Signed-off-by:
-
Doug Ledford authored
commit f4e7de63 upstream. When we fail to find the default gid index, we can't continue processing in this routine or else we will pass a negative index to later routines resulting in invalid memory access attempts and a kernel oops. Fixes: 03db3a2d (IB/core: Add RoCE GID table management) Signed-off-by:
-
Michael Neuling authored
commit d6776bba upstream. Keep IRQ mappings on context teardown. This won't leak IRQs as if we allocate the mapping again, the generic code will give the same mapping used last time. Doing this works around a race in the generic code. Masking the interrupt introduces a race which can crash the kernel or result in IRQ that is never EOIed. The lost of EOI results in all subsequent mappings to the same HW IRQ never receiving an interrupt. We've seen this race with cxl test cases which are doing heavy context startup and teardown at the same time as heavy interrupt load. A fix to the generic code is being investigated also. Signed-off-by:
Michael Neuling <mikey@neuling.org> Tested-by:
Andrew Donnellan <andrew.donnellan@au1.ibm.com> Acked-by:
Ian Munsie <imunsie@au1.ibm.com> Tested-by:
Vaibhav Jain <vaibhav@linux.vnet.ibm.com> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Signed-off-by:
-
Hans Verkuil authored
commit 3020ca71 upstream. The VSync polarity was negative instead of positive for the 4k CEA formats. I probably copy-and-pasted these from the DMT 4k format, which does have a negative VSync polarity. Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Reported-by:
Martin Bugge <marbugge@cisco.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@osg.samsung.com> Signed-off-by:
-
Sakari Ailus authored
commit 2c1f6951 upstream. When a buffer is being dequeued using VIDIOC_DQBUF IOCTL, the exact buffer which will be dequeued is not known until the buffer has been removed from the queue. The number of planes is specific to a buffer, not to the queue. This does lead to the situation where multi-plane buffers may be requested and queued with n planes, but VIDIOC_DQBUF IOCTL may be passed an argument struct with fewer planes. __fill_v4l2_buffer() however uses the number of planes from the dequeued videobuf2 buffer, overwriting kernel memory (the m.planes array allocated in video_usercopy() in v4l2-ioctl.c) if the user provided fewer planes than the dequeued buffer had. Oops! Fixes: b0e0e1f8 ("[media] media: videobuf2: Prepare to divide videobuf2") Signed-off-by:
Sakari Ailus <sakari.ailus@linux.intel.com> Acked-by:
-
Sakari Ailus authored
commit e7e0c3e2 upstream. The number of planes in videobuf2 is specific to a buffer. In order to verify that the planes array provided by the user is long enough, a new vb2_buf_op is required. Call __verify_planes_array() when the dequeued buffer is known. Return an error to the caller if there was one, otherwise remove the buffer from the done list. Signed-off-by:
-
Ricardo Ribalda Delgado authored
commit b9387684 upstream. When using a device is read/write mode, vb2 does not handle properly the first select/poll operation. The reason for this, is that when this code has been refactored, some of the operations have changed their order, and now fileio emulator is not started. The reintroduced check to the core is enabled by a quirk flag, that avoids this check by other subsystems like DVB. Fixes: 49d8ab9f ("media] media: videobuf2: Separate vb2_poll()") Reported-by:
Dimitrios Katsaros <patcherwork@gmail.com> Cc: Junghak Sung <jh1009.sung@samsung.com> Signed-off-by:
Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com> Signed-off-by:
-
Ricardo Ribalda Delgado authored
commit 89a09566 upstream. On page unaligned frames, create_framevec forces get_vaddr_frames to allocate an extra page at the end of the buffer. Under some circumstances, this leads to -EINVAL on VIDIOC_QBUF. E.g: We have vm_a that vm_area that goes from 0x1000 to 0x3000. And a frame that goes from 0x1800 to 0x2800, i.e. 2 pages. frame_vector_create will be called with the following params: get_vaddr_frames(0x1800, 2, write, 1, vec); get_vaddr will allocate the first page after checking that the memory 0x1800-0x27ff is valid, but it will not allocate the second page because the range 0x2800-0x37ff is out of the vm_a range. This results in create_framevec returning -EFAULT Error Trace: [ 9083.793015] video0: VIDIOC_QBUF: 00:00:00.00000000 index=1, type=vid-cap, flags=0x00002002, field=any, sequence=0, memory=userptr, bytesused=0, offset/userptr=0x7ff2b023ca80, length=5765760 [ 9083.793028] timecode=00:00:00 type=0, flags=0x00000000, frames=0, userbits=0x00000000 [ 9083.793117] video0: VIDIOC_QBUF: error -22: 00:00:00.00000000 index=2, type=vid-cap, flags=0x00000000, field=any, sequence=0, memory=userptr, bytesused=0, offset/userptr=0x7ff2b07bc500, length=5765760 Also use true instead of 1 since that argument is a bool in the get_vaddr_frames() prototype. Fixes: 21fb0cb7 ("[media] vb2: Provide helpers for mapping virtual addresses") Reported-by:
Albert Antony <albert@newtec.dk> Signed-off-by:
Marek Szyprowski <m.szyprowski@samsung.com> Reviewed-by:
Jan Kara <jack@suse.cz> Signed-off-by:
-
Sugar Zhang authored
commit 653aa464 upstream. this patch corrects the interface adc/dac control register definition according to datasheet. Signed-off-by:
Sugar Zhang <sugar.zhang@rock-chips.com> Signed-off-by:
Mark Brown <broonie@kernel.org> Signed-off-by:
-
Mark Brown authored
commit 47325078 upstream. The dummy component is reused for all cards so we special case and don't bind it to any of them. This means that code like that displaying the component widgets that tries to look at the card will crash. In the future we will fix this by ensuring that the dummy component looks like other components but that is invasive and so not suitable for a fix. Instead add a special case check here. Reported-by:
Harry Pan <harry.pan@intel.com> Suggested-by:
Lars-Peter Clausen <lars@metafoo.de> Signed-off-by:
-
Lars-Peter Clausen authored
commit 712a8038 upstream. When the ssm4567 is powered up the driver calles regcache_sync() to restore the register map content. regcache_sync() assumes that the device is in its power-on reset state. Make sure that this is the case by explicitly resetting the ssm4567 register map before calling regcache_sync() otherwise we might end up with a incorrect register map which leads to undefined behaviour. One such undefined behaviour was observed when returning from system suspend while a playback stream is active, in that case the ssm4567 was kept muted after resume. Fixes: 1ee44ce0 ("ASoC: ssm4567: Add driver for Analog Devices SSM4567 amplifier") Reported-by:
Harsha Priya <harshapriya.n@intel.com> Tested-by:
Fang, Yang A <yang.a.fang@intel.com> Signed-off-by:
-
Arnd Bergmann authored
commit ba4bc32e upstream. An older patch to convert the API in the s3c i2s driver ended up passing a const pointer into a function that takes a non-const pointer, so we now get a warning: sound/soc/samsung/s3c2412-i2s.c: In function 's3c2412_iis_dev_probe': sound/soc/samsung/s3c2412-i2s.c:172:9: error: passing argument 3 of 's3c_i2sv2_register_component' discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers] However, the s3c_i2sv2_register_component() function again passes the pointer into another function taking a const, so we just need to change its prototype. Fixes: eca3b01d ("ASoC: switch over to use snd_soc_register_component() on s3c i2s") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Reviewed-by:
Krzysztof Kozlowski <k.kozlowski@samsung.com> Signed-off-by:
-
Tony Luck authored
commit c4fc1956 upstream. Both of these drivers can return NOTIFY_BAD, but this terminates processing other callbacks that were registered later on the chain. Since the driver did nothing to log the error it seems wrong to prevent other interested parties from seeing it. E.g. neither of them had even bothered to check the type of the error to see if it was a memory error before the return NOTIFY_BAD. Signed-off-by:
Tony Luck <tony.luck@intel.com> Acked-by:
Aristeu Rozanski <aris@redhat.com> Acked-by:
Borislav Petkov <bp@suse.de> Signed-off-by:
-
Azael Avalos authored
commit a30b8f81 upstream. Commit 52cbae01 ("toshiba_acpi: Change default Hotkey enabling value") changed the hotkeys enabling value, as it was the same value Windows uses, however, it turns out that the value tells the EC that the driver will now take care of the hardware events like the physical RFKill switch or the pointing device toggle button. This patch reverts such commit by changing the default hotkey enabling value to 0x09, which enables hotkey events only, making the hardware buttons working again. Fixes bugs 113331 and 114941. Signed-off-by:
Azael Avalos <coproscefalo@gmail.com> Signed-off-by:
Darren Hart <dvhart@linux.intel.com> Signed-off-by:
-
Adrian Hunter authored
commit 6e1c7d61 upstream. Baytrail eMMC/SD/SDIO host controllers have been known to hang. A change to a hardware setting has been found to reduce the occurrence of such hangs. This patch ensures the correct setting. This patch applies cleanly to v4.4+. It could go to earlier kernels also, so I will send backports to the stable list in due course. Signed-off-by:
Adrian Hunter <adrian.hunter@intel.com> Signed-off-by:
Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by:
-
Lv Zheng authored
commit 5508df89 upstream. It is reported that the following commit triggers regressions: Linux commit: efaed9be ACPICA commit: 31178590dde82368fdb0f6b0e466b6c0add96c57 Subject: ACPICA: Events: Enhance acpi_ev_execute_reg_method() to ensure no _REG evaluations can happen during OS early boot stages This is because that the ECDT support is not corrected in Linux, and Linux requires to execute _REG for ECDT (though this sounds so wrong), we need to ensure acpi_gbl_namespace_initialized is set before ECDT probing in order for _REG to be executed. Since we have to move "acpi_gbl_namespace_initialized = TRUE" to the initialization step happening before ECDT probing, acpi_load_tables() is the best candidate for now. Thus this patch fixes the regression by doing so. But if the ECDT support is fixed, Linux will not execute _REG for ECDT, and ECDT probing will happen before acpi_load_tables(). At that time, we still want to ensure acpi_gbl_namespace_initialized is set after executing acpi_ns_initialize_objects() (under the condition of acpi_gbl_group_module_level_code = FALSE), this patch also moves acpi_ns_initialize_objects() to acpi_load_tables() accordingly. Since acpi_ns_initialize_objects() doesn't seem to be skippable, this patch also removes ACPI_NO_OBJECT_INIT for the one invoked in acpi_load_tables(). And since the default region handlers should always be installed before loading the tables, this patch also removes useless acpi_gbl_group_module_level_code check accordingly. Reported by Chris Bainbridge, Fixed by Lv Zheng. Fixes: efaed9be (ACPICA: Events: Enhance acpi_ev_execute_reg_method() to ensure no _REG evaluations can happen during OS early boot stages) Reported-and-tested-by:
Chris Bainbridge <chris.bainbridge@gmail.com> Signed-off-by:
Lv Zheng <lv.zheng@intel.com> Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by:
-
Javier Martinez Canillas authored
commit 10ff4c52 upstream. The exynos5 I2C controller driver always prepares and enables a clock before using it and then disables unprepares it when the clock is not used anymore. But this can cause a possible ABBA deadlock in some scenarios since a driver that uses regmap to access its I2C registers, will first grab the regmap lock and then the I2C xfer function will grab the prepare lock when preparing the I2C clock. But since the clock driver also uses regmap for I2C accesses, preparing a clock will first grab the prepare lock and then the regmap lock when using the regmap API. An example of this happens on the Exynos5422 Odroid XU4 board where a s2mps11 PMIC is used and both the s2mps11 regulators and clk drivers share the same I2C regmap. The possible deadlock is reported by the kernel lockdep: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(sec_core:428:(regmap)->lock); lock(prepare_lock); lock(sec_core:428:(regmap)->lock); lock(prepare_lock); *** DEADLOCK *** Fix it by leaving the code prepared on probe and use {en,dis}able in the I2C transfer function. This patch is similar to commit 34e81ad5 ("i2c: s3c2410: fix ABBA deadlock by keeping clock prepared") that fixes the same bug in other driver for an I2C controller found in Samsung SoCs. Reported-by:
Anand Moon <linux.amoon@gmail.com> Signed-off-by:
Javier Martinez Canillas <javier@osg.samsung.com> Reviewed-by:
Wolfram Sang <wsa@the-dreams.de> Signed-off-by:
-
Michael Ellerman authored
commit 609d5a1b upstream. Since commit ea8daa7b ("kbuild: Add option to turn incompatible pointer check into error"), assignments from an incompatible pointer types have become a hard error, eg: drivers/i2c/busses/i2c-cpm.c:545:91: error: passing argument 3 of 'dma_alloc_coherent' from incompatible pointer type Fix the build break by converting txdma & rxdma to dma_addr_t. Signed-off-by:
-
Adrian Hunter authored
commit 1342e0b7 upstream. Tracing a workload that uses transactions gave a seg fault as follows: perf record -e intel_pt// workload perf report Program received signal SIGSEGV, Segmentation fault. 0x000000000054b58c in intel_pt_reset_last_branch_rb (ptq=0x1a36110) at util/intel-pt.c:929 929 ptq->last_branch_rb->nr = 0; (gdb) p ptq->last_branch_rb $1 = (struct branch_stack *) 0x0 (gdb) up 1148 intel_pt_reset_last_branch_rb(ptq); (gdb) l 1143 if (ret) 1144 pr_err("Intel Processor Trace: failed to deliver transaction event 1145 ret); 1146 1147 if (pt->synth_opts.callchain) 1148 intel_pt_reset_last_branch_rb(ptq); 1149 1150 return ret; 1151 } 1152 (gdb) p pt->synth_opts.callchain $2 = true (gdb) (gdb) bt #0 0x000000000054b58c in intel_pt_reset_last_branch_rb (ptq=0x1a36110) #1 0x000000000054c1e0 in intel_pt_synth_transaction_sample (ptq=0x1a36110) #2 0x000000000054c5b2 in intel_pt_sample (ptq=0x1a36110) Caused by checking the 'callchain' flag when it should have been the 'last_branch' flag. Fix that. Reported-by:
Andi Kleen <ak@linux.intel.com> Signed-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by:
-
Linus Walleij authored
commit f36fdacc upstream. The current compile-time check for inversed IENB/CNTL does not work in multiplatform boots: as soon as versatile is included in the build, the IENB/CNTL is switched and breaks graphics. Convert this to a runtime switch. Cc: Rob Herring <robh@kernel.org> Cc: Russell King <linux@arm.linux.org.uk> Fixes: a29da136 ("ARM: versatile: convert to multi-platform") Signed-off-by:
Linus Walleij <linus.walleij@linaro.org> Signed-off-by:
Tomi Valkeinen <tomi.valkeinen@ti.com> Signed-off-by:
-
Flora Cui authored
commit 56fc3502 upstream. Fixes the following scenario: 1. Page table bo allocated in vram and linked to man->lru. tbo->list_kref.refcount=2 2. Page table bo is swapped out and removed from man->lru. tbo->list_kref.refcount=1 3. Command submission from userspace. Page table bo is moved to vram. ttm_bo_move_to_lru_tail() link it to man->lru and don't increase the kref count. Reviewed-by:
Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by:
Flora Cui <Flora.Cui@amd.com> Reviewed-by:
Christian König <christian.koenig@amd.com> Signed-off-by:
Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
-
Ville Syrjälä authored
commit 31318a92 upstream. HSW still has the wake FIFO, so let's check it. Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Deepak S <deepak.s@linux.intel.com> Fixes: 05a2fb15 ("drm/i915: Consolidate forcewake code") Signed-off-by:
Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1460633942-24013-1-git-send-email-ville.syrjala@linux.intel.comReviewed-by:
Mika Kuoppala <mika.kuoppala@intel.com> (cherry picked from commit 3d7d0c85) Signed-off-by:
Jani Nikula <jani.nikula@intel.com> Signed-off-by:
-
Chris Wilson authored
commit 1b3e885a upstream. For reasons unknown Sandybridge GT1 (at least) will eventually hang when it encounters a ring wraparound at offset 0. The test case that reproduces the bug reliably forces a large number of interrupted context switches, thereby causing very frequent ring wraparounds, but there are similar bug reports in the wild with the same symptoms, seqno writes stop just before the wrap and the ringbuffer at address 0. It is also timing crucial, but adding various delays hasn't helped pinpoint where the window lies. Whether the fault is restricted to the ringbuffer itself or the GTT addressing is unclear, but moving the ringbuffer fixes all the hangs I have been able to reproduce. References: (e.g.) https://bugs.freedesktop.org/show_bug.cgi?id=93262 Testcase: igt/gem_exec_whisper/render-contexts-interruptible #snb-gt1 Signed-off-by:
Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/1460565315-7748-12-git-send-email-chris@chris-wilson.co.uk (cherry picked from commit a687a43a) Signed-off-by:
-
Akash Goel authored
commit d43f3ebf upstream. Currently for the case where there is enough space at the end of Ring buffer for accommodating only the base request, the wrapround is done immediately and as a result the base request gets added at the start of Ring buffer. But there may not be enough free space at the beginning to accommodate the base request, as before the wraparound, the wait was effectively done for the reserved_size free space from the start of Ring buffer. In such a case there is a potential of Ring buffer overflow, the instructions at the head of Ring (ACTHD) can get overwritten. Since the base request can fit in the remaining space, there is no need to wraparound immediately. The wraparound will anyway happen later when the reserved part starts getting used. Cc: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by:
Akash Goel <akash.goel@intel.com> Signed-off-by:
-
Mika Kuoppala authored
commit 510650e8 upstream. Experiments with heaven 4.0 benchmark and skylake gt3e (rev 0xa) suggest that WaForceContextSaveRestoreNonCoherent is needed for all revs. Extending this to all revs cures a gpu hang with rev 0xa when running heaven4.0 gpu benchmark. We have been here before, with problems enabling gt4e and extending up to revision F0 instead of false claims of bspec of E0 only. See commit <e238659d> ("drm/i915/skl: Default to noncoherent access up to F0"). In retrospect we should have covered this with this big blanket back then already, as E0 vs F0 discrepancy was suspicious enough. Previously the WaForceEnableNonCoherent has been tied to context non-coherence, atleast in relevant hsds. So keep this tie and extended this alongside. Cc: Abdiel Janulgue <abdiel.janulgue@linux.intel.com> Cc: Ben Widawsky <benjamin.widawsky@intel.com> Cc: Timo Aaltonen <tjaalton@ubuntu.com> Reported-by:
Mike Lothian <mike@fireburn.co.uk> References: https://bugs.freedesktop.org/show_bug.cgi?id=93491Signed-off-by:
Ben Widawsky <benjamin.widawsky@intel.com> Tested-by:
Timo Aaltonen <tjaalton@ubuntu.com> Link: http://patchwork.freedesktop.org/patch/msgid/1459860977-27751-2-git-send-email-mika.kuoppala@intel.com (cherry picked from commit 97ea6be1) Signed-off-by:
-
Mat Martineau authored
commit a41c8882 upstream. The driver does not load firmware for unknown steppings, so these new steppings must be added to the list. Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by:
Mat Martineau <mathew.j.martineau@linux.intel.com> Reviewed-by:
Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by:
-
Dan Carpenter authored
commit 93fce954 upstream. At the end of the function we expect "status" to be zero, but it's either -EINVAL or uninitialized. Fixes: 788bf83d ('drm/amdkfd: Add wave control operation to debugger') Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
Oded Gabbay <oded.gabbay@gmail.com> Signed-off-by:
-
Ville Syrjälä authored
commit fa5a7970 upstream. Pass BIT(DRM_ROTATE_0) instead of DRM_ROTATE_0 to skl_update_scaler(). The former is a mask, the latter just the bit number. Fortunately the only thing skl_update_scaler() does with the rotation is check if it's 90/270 degrees or not, and so in this case it would still do the right thing. Cc: Chandra Konduru <chandra.konduru@intel.com> Signed-off-by:
Matt Roper <matthew.d.roper@intel.com> Signed-off-by:
-
Ville Syrjälä authored
commit f1ecaf8f upstream. ilk_compute_pipe_wm() assumes as zeroed pipe_wm structure when it starts. We used to pass such a zeroed struct in, but this got broken when the pipe_wm structure got embedded in the crtc state. To fix it without too much fuzz, we need to resort to a memset(). Fixes: 4e0963c7 ("drm/i915: Calculate pipe watermarks into CRTC state (v3)") Cc: Matt Roper <matthew.d.roper@intel.com> Signed-off-by:
-
Ville Syrjälä authored
commit d890565c upstream. ilk_program_watermarks() is supposed to merge the active watermarks from all pipes. Thus we need to use the active config too instead of some precomputed stuff. Fixes: aa363136 ("drm/i915: Calculate watermark configuration during atomic check (v2)") Cc: Matt Roper <matthew.d.roper@intel.com> Signed-off-by:
-
