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

[Linux-stable-mirror] Patch "drm/amdkfd: Fix SDMA ring buffer size calculation" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdkfd: Fix SDMA ring buffer size calculation to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdkfd-fix-sdma-ring-buffer-size-calculation.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: shaoyunl <Shaoyun.Liu(a)amd.com> Date: Wed, 1 Nov 2017 19:21:56 -0400 Subject: drm/amdkfd: Fix SDMA ring buffer size calculation From: shaoyunl <Shaoyun.Liu(a)amd.com> [ Upstream commit d12fb13f23199faa7e536acec1db49068e5a067d ] ffs function return the position of the first bit set on 1 based. (bit zero returns 1). Signed-off-by: shaoyun liu <shaoyun.liu(a)amd.com> Signed-off-by: Felix Kuehling <Felix.Kuehling(a)amd.com> Reviewed-by: Oded Gabbay <oded.gabbay(a)gmail.com> Signed-off-by: Oded Gabbay <oded.gabbay(a)gmail.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c @@ -205,8 +205,8 @@ static int update_mqd_sdma(struct mqd_ma struct cik_sdma_rlc_registers *m; m = get_sdma_mqd(mqd); - m->sdma_rlc_rb_cntl = ffs(q->queue_size / sizeof(unsigned int)) << - SDMA0_RLC0_RB_CNTL__RB_SIZE__SHIFT | + m->sdma_rlc_rb_cntl = (ffs(q->queue_size / sizeof(unsigned int)) - 1) + << SDMA0_RLC0_RB_CNTL__RB_SIZE__SHIFT | q->vmid << SDMA0_RLC0_RB_CNTL__RB_VMID__SHIFT | 1 << SDMA0_RLC0_RB_CNTL__RPTR_WRITEBACK_ENABLE__SHIFT | 6 << SDMA0_RLC0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT; Patches currently in stable-queue which might be from Shaoyun.Liu(a)amd.com are queue-4.14/drm-amdkfd-fix-sdma-ring-buffer-size-calculation.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "drm/amdkfd: Fix SDMA oversubsription handling" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdkfd: Fix SDMA oversubsription handling to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdkfd-fix-sdma-oversubsription-handling.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: Felix Kuehling <Felix.Kuehling(a)amd.com> Date: Wed, 1 Nov 2017 19:21:57 -0400 Subject: drm/amdkfd: Fix SDMA oversubsription handling From: Felix Kuehling <Felix.Kuehling(a)amd.com> [ Upstream commit 8c946b8988acec785bcf67088b6bd0747f36d2d3 ] SDMA only supports a fixed number of queues. HWS cannot handle oversubscription. Signed-off-by: shaoyun liu <shaoyun.liu(a)amd.com> Signed-off-by: Felix Kuehling <Felix.Kuehling(a)amd.com> Reviewed-by: Oded Gabbay <oded.gabbay(a)gmail.com> Signed-off-by: Oded Gabbay <oded.gabbay(a)gmail.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c | 18 +++++++++++++++++ 1 file changed, 18 insertions(+) --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -184,6 +184,24 @@ int pqm_create_queue(struct process_queu switch (type) { case KFD_QUEUE_TYPE_SDMA: + if (dev->dqm->queue_count >= + CIK_SDMA_QUEUES_PER_ENGINE * CIK_SDMA_ENGINE_NUM) { + pr_err("Over-subscription is not allowed for SDMA.\n"); + retval = -EPERM; + goto err_create_queue; + } + + retval = create_cp_queue(pqm, dev, &q, properties, f, *qid); + if (retval != 0) + goto err_create_queue; + pqn->q = q; + pqn->kq = NULL; + retval = dev->dqm->ops.create_queue(dev->dqm, q, &pdd->qpd, + &q->properties.vmid); + pr_debug("DQM returned %d for create_queue\n", retval); + print_queue(q); + break; + case KFD_QUEUE_TYPE_COMPUTE: /* check if there is over subscription */ if ((sched_policy == KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) && Patches currently in stable-queue which might be from Felix.Kuehling(a)amd.com are queue-4.14/drm-amdkfd-fix-sdma-oversubsription-handling.patch queue-4.14/drm-amdkfd-fix-sdma-ring-buffer-size-calculation.patch queue-4.14/drm-amdgpu-fix-sdma-load-unload-sequence-on-hws-disabled-mode.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "drm/amdgpu: Fix SDMA load/unload sequence on HWS disabled mode" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdgpu: Fix SDMA load/unload sequence on HWS disabled mode to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdgpu-fix-sdma-load-unload-sequence-on-hws-disabled-mode.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: Felix Kuehling <Felix.Kuehling(a)amd.com> Date: Wed, 1 Nov 2017 19:21:55 -0400 Subject: drm/amdgpu: Fix SDMA load/unload sequence on HWS disabled mode From: Felix Kuehling <Felix.Kuehling(a)amd.com> [ Upstream commit cf21654b40968609779751b34e7923180968fe5b ] Fix the SDMA load and unload sequence as suggested by HW document. Signed-off-by: shaoyun liu <shaoyun.liu(a)amd.com> Signed-off-by: Felix Kuehling <Felix.Kuehling(a)amd.com> Acked-by: Oded Gabbay <oded.gabbay(a)gmail.com> Signed-off-by: Oded Gabbay <oded.gabbay(a)gmail.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c | 47 +++++++++++++++------- 1 file changed, 34 insertions(+), 13 deletions(-) --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c @@ -369,29 +369,50 @@ static int kgd_hqd_sdma_load(struct kgd_ { struct amdgpu_device *adev = get_amdgpu_device(kgd); struct cik_sdma_rlc_registers *m; + unsigned long end_jiffies; uint32_t sdma_base_addr; + uint32_t data; m = get_sdma_mqd(mqd); sdma_base_addr = get_sdma_base_addr(m); - WREG32(sdma_base_addr + mmSDMA0_RLC0_VIRTUAL_ADDR, - m->sdma_rlc_virtual_addr); + WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, + m->sdma_rlc_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)); - WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, - m->sdma_rlc_rb_base); + end_jiffies = msecs_to_jiffies(2000) + jiffies; + while (true) { + data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS); + if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK) + break; + if (time_after(jiffies, end_jiffies)) + return -ETIME; + usleep_range(500, 1000); + } + if (m->sdma_engine_id) { + data = RREG32(mmSDMA1_GFX_CONTEXT_CNTL); + data = REG_SET_FIELD(data, SDMA1_GFX_CONTEXT_CNTL, + RESUME_CTX, 0); + WREG32(mmSDMA1_GFX_CONTEXT_CNTL, data); + } else { + data = RREG32(mmSDMA0_GFX_CONTEXT_CNTL); + data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL, + RESUME_CTX, 0); + WREG32(mmSDMA0_GFX_CONTEXT_CNTL, data); + } + WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, + m->sdma_rlc_doorbell); + WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, 0); + WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR, 0); + WREG32(sdma_base_addr + mmSDMA0_RLC0_VIRTUAL_ADDR, + m->sdma_rlc_virtual_addr); + WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdma_rlc_rb_base); WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI, m->sdma_rlc_rb_base_hi); - WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO, m->sdma_rlc_rb_rptr_addr_lo); - WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI, m->sdma_rlc_rb_rptr_addr_hi); - - WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, - m->sdma_rlc_doorbell); - WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, m->sdma_rlc_rb_cntl); @@ -564,9 +585,9 @@ static int kgd_hqd_sdma_destroy(struct k } WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0); - WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, 0); - WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR, 0); - WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, 0); + WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, + RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) | + SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK); return 0; } Patches currently in stable-queue which might be from Felix.Kuehling(a)amd.com are queue-4.14/drm-amdkfd-fix-sdma-oversubsription-handling.patch queue-4.14/drm-amdkfd-fix-sdma-ring-buffer-size-calculation.patch queue-4.14/drm-amdgpu-fix-sdma-load-unload-sequence-on-hws-disabled-mode.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "drm/amdgpu: don't try to move pinned BOs" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdgpu: don't try to move pinned BOs to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdgpu-don-t-try-to-move-pinned-bos.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: "Christian König" <christian.koenig(a)amd.com> Date: Fri, 24 Nov 2017 11:39:30 +0100 Subject: drm/amdgpu: don't try to move pinned BOs From: "Christian König" <christian.koenig(a)amd.com> [ Upstream commit 6edc6910ba4cd6eab309263539c8f09b8ad772bf ] Never try to move pinned BOs during CS. Signed-off-by: Christian KÃ¶nig <christian.koenig(a)amd.com> Reviewed-by: Michel DÃ¤nzer <michel.daenzer(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 4 ++++ 1 file changed, 4 insertions(+) --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -403,6 +403,10 @@ static bool amdgpu_cs_try_evict(struct a if (candidate->robj == validated) break; + /* We can't move pinned BOs here */ + if (bo->pin_count) + continue; + other = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type); /* Check if this BO is in one of the domains we need space for */ Patches currently in stable-queue which might be from christian.koenig(a)amd.com are queue-4.14/drm-amdgpu-don-t-try-to-move-pinned-bos.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "cpupowerutils: bench - Fix cpu online check" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled cpupowerutils: bench - Fix cpu online check to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: cpupowerutils-bench-fix-cpu-online-check.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: Abhishek Goel <huntbag(a)linux.vnet.ibm.com> Date: Tue, 7 Nov 2017 15:17:55 +0530 Subject: cpupowerutils: bench - Fix cpu online check From: Abhishek Goel <huntbag(a)linux.vnet.ibm.com> [ Upstream commit 53d1cd6b125fb9d69303516a1179ebc3b72f797a ] cpupower_is_cpu_online was incorrectly checking for 0. This patch fixes this by checking for 1 when the cpu is online. Signed-off-by: Abhishek Goel <huntbag(a)linux.vnet.ibm.com> Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- tools/power/cpupower/bench/system.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/tools/power/cpupower/bench/system.c +++ b/tools/power/cpupower/bench/system.c @@ -61,7 +61,7 @@ int set_cpufreq_governor(char *governor, dprintf("set %s as cpufreq governor\n", governor); - if (cpupower_is_cpu_online(cpu) != 0) { + if (cpupower_is_cpu_online(cpu) != 1) { perror("cpufreq_cpu_exists"); fprintf(stderr, "error: cpu %u does not exist\n", cpu); return -1; Patches currently in stable-queue which might be from huntbag(a)linux.vnet.ibm.com are queue-4.14/cpupowerutils-bench-fix-cpu-online-check.patch queue-4.14/cpupower-fix-cpupower-working-when-cpu0-is-offline.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "cpupower : Fix cpupower working when cpu0 is offline" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled cpupower : Fix cpupower working when cpu0 is offline to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: cpupower-fix-cpupower-working-when-cpu0-is-offline.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: Abhishek Goel <huntbag(a)linux.vnet.ibm.com> Date: Wed, 15 Nov 2017 14:10:02 +0530 Subject: cpupower : Fix cpupower working when cpu0 is offline From: Abhishek Goel <huntbag(a)linux.vnet.ibm.com> [ Upstream commit dbdc468f35ee827cab2753caa1c660bdb832243a ] cpuidle_monitor used to assume that cpu0 is always online which is not a valid assumption on POWER machines. This patch fixes this by getting the cpu on which the current thread is running, instead of always using cpu0 for monitoring which may not be online. Signed-off-by: Abhishek Goel <huntbag(a)linux.vnet.ibm.com> Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- tools/power/cpupower/utils/idle_monitor/cpuidle_sysfs.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) --- a/tools/power/cpupower/utils/idle_monitor/cpuidle_sysfs.c +++ b/tools/power/cpupower/utils/idle_monitor/cpuidle_sysfs.c @@ -130,15 +130,18 @@ static struct cpuidle_monitor *cpuidle_r { int num; char *tmp; + int this_cpu; + + this_cpu = sched_getcpu(); /* Assume idle state count is the same for all CPUs */ - cpuidle_sysfs_monitor.hw_states_num = cpuidle_state_count(0); + cpuidle_sysfs_monitor.hw_states_num = cpuidle_state_count(this_cpu); if (cpuidle_sysfs_monitor.hw_states_num <= 0) return NULL; for (num = 0; num < cpuidle_sysfs_monitor.hw_states_num; num++) { - tmp = cpuidle_state_name(0, num); + tmp = cpuidle_state_name(this_cpu, num); if (tmp == NULL) continue; @@ -146,7 +149,7 @@ static struct cpuidle_monitor *cpuidle_r strncpy(cpuidle_cstates[num].name, tmp, CSTATE_NAME_LEN - 1); free(tmp); - tmp = cpuidle_state_desc(0, num); + tmp = cpuidle_state_desc(this_cpu, num); if (tmp == NULL) continue; strncpy(cpuidle_cstates[num].desc, tmp, CSTATE_DESC_LEN - 1); Patches currently in stable-queue which might be from huntbag(a)linux.vnet.ibm.com are queue-4.14/cpupowerutils-bench-fix-cpu-online-check.patch queue-4.14/cpupower-fix-cpupower-working-when-cpu0-is-offline.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "cpufreq: Add Loongson machine dependencies" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled cpufreq: Add Loongson machine dependencies to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: cpufreq-add-loongson-machine-dependencies.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: James Hogan <jhogan(a)kernel.org> Date: Wed, 15 Nov 2017 21:17:55 +0000 Subject: cpufreq: Add Loongson machine dependencies From: James Hogan <jhogan(a)kernel.org> [ Upstream commit 0d307935fefa6389eb726c6362351c162c949101 ] The MIPS loongson cpufreq drivers don't build unless configured for the correct machine type, due to dependency on machine specific architecture headers and symbols in machine specific platform code. More specifically loongson1-cpufreq.c uses RST_CPU_EN and RST_CPU, neither of which is defined in asm/mach-loongson32/regs-clk.h unless CONFIG_LOONGSON1_LS1B=y, and loongson2_cpufreq.c references loongson2_clockmod_table[], which is only defined in arch/mips/loongson64/lemote-2f/clock.c, i.e. when CONFIG_LEMOTE_MACH2F=y. Add these dependencies to Kconfig to avoid randconfig / allyesconfig build failures (e.g. when based on BMIPS which also has a cpufreq driver). Signed-off-by: James Hogan <jhogan(a)kernel.org> Acked-by: Viresh Kumar <viresh.kumar(a)linaro.org> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/cpufreq/Kconfig | 2 ++ 1 file changed, 2 insertions(+) --- a/drivers/cpufreq/Kconfig +++ b/drivers/cpufreq/Kconfig @@ -275,6 +275,7 @@ config BMIPS_CPUFREQ config LOONGSON2_CPUFREQ tristate "Loongson2 CPUFreq Driver" + depends on LEMOTE_MACH2F help This option adds a CPUFreq driver for loongson processors which support software configurable cpu frequency. @@ -287,6 +288,7 @@ config LOONGSON2_CPUFREQ config LOONGSON1_CPUFREQ tristate "Loongson1 CPUFreq Driver" + depends on LOONGSON1_LS1B help This option adds a CPUFreq driver for loongson1 processors which support software configurable cpu frequency. Patches currently in stable-queue which might be from jhogan(a)kernel.org are queue-4.14/cpufreq-add-loongson-machine-dependencies.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "Btrfs: incremental send, fix wrong unlink path after renaming file" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled Btrfs: incremental send, fix wrong unlink path after renaming file to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: btrfs-incremental-send-fix-wrong-unlink-path-after-renaming-file.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: Filipe Manana <fdmanana(a)suse.com> Date: Fri, 17 Nov 2017 01:54:00 +0000 Subject: Btrfs: incremental send, fix wrong unlink path after renaming file From: Filipe Manana <fdmanana(a)suse.com> [ Upstream commit ea37d5998b50a72b9045ba60a132eeb20e1c4230 ] Under some circumstances, an incremental send operation can issue wrong paths for unlink commands related to files that have multiple hard links and some (or all) of those links were renamed between the parent and send snapshots. Consider the following example: Parent snapshot . (ino 256) |---- a/ (ino 257) | |---- b/ (ino 259) | | |---- c/ (ino 260) | | |---- f2 (ino 261) | | | |---- f2l1 (ino 261) | |---- d/ (ino 262) |---- f1l1_2 (ino 258) |---- f2l2 (ino 261) |---- f1_2 (ino 258) Send snapshot . (ino 256) |---- a/ (ino 257) | |---- f2l1/ (ino 263) | |---- b2/ (ino 259) | |---- c/ (ino 260) | | |---- d3 (ino 262) | | |---- f1l1_2 (ino 258) | | |---- f2l2_2 (ino 261) | | |---- f1_2 (ino 258) | | | |---- f2 (ino 261) | |---- f1l2 (ino 258) | |---- d (ino 261) When computing the incremental send stream the following steps happen: 1) When processing inode 261, a rename operation is issued that renames inode 262, which currently as a path of "d", to an orphan name of "o262-7-0". This is done because in the send snapshot, inode 261 has of its hard links with a path of "d" as well. 2) Two link operations are issued that create the new hard links for inode 261, whose names are "d" and "f2l2_2", at paths "/" and "o262-7-0/" respectively. 3) Still while processing inode 261, unlink operations are issued to remove the old hard links of inode 261, with names "f2l1" and "f2l2", at paths "a/" and "d/". However path "d/" does not correspond anymore to the directory inode 262 but corresponds instead to a hard link of inode 261 (link command issued in the previous step). This makes the receiver fail with a ENOTDIR error when attempting the unlink operation. The problem happens because before sending the unlink operation, we failed to detect that inode 262 was one of ancestors for inode 261 in the parent snapshot, and therefore we didn't recompute the path for inode 262 before issuing the unlink operation for the link named "f2l2" of inode 262. The detection failed because the function "is_ancestor()" only follows the first hard link it finds for an inode instead of all of its hard links (as it was originally created for being used with directories only, for which only one hard link exists). So fix this by making "is_ancestor()" follow all hard links of the input inode. A test case for fstests follows soon. Signed-off-by: Filipe Manana <fdmanana(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- fs/btrfs/send.c | 124 +++++++++++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 106 insertions(+), 18 deletions(-) --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -3527,7 +3527,40 @@ out: } /* - * Check if ino ino1 is an ancestor of inode ino2 in the given root. + * Check if inode ino2, or any of its ancestors, is inode ino1. + * Return 1 if true, 0 if false and < 0 on error. + */ +static int check_ino_in_path(struct btrfs_root *root, + const u64 ino1, + const u64 ino1_gen, + const u64 ino2, + const u64 ino2_gen, + struct fs_path *fs_path) +{ + u64 ino = ino2; + + if (ino1 == ino2) + return ino1_gen == ino2_gen; + + while (ino > BTRFS_FIRST_FREE_OBJECTID) { + u64 parent; + u64 parent_gen; + int ret; + + fs_path_reset(fs_path); + ret = get_first_ref(root, ino, &parent, &parent_gen, fs_path); + if (ret < 0) + return ret; + if (parent == ino1) + return parent_gen == ino1_gen; + ino = parent; + } + return 0; +} + +/* + * Check if ino ino1 is an ancestor of inode ino2 in the given root for any + * possible path (in case ino2 is not a directory and has multiple hard links). * Return 1 if true, 0 if false and < 0 on error. */ static int is_ancestor(struct btrfs_root *root, @@ -3536,36 +3569,91 @@ static int is_ancestor(struct btrfs_root const u64 ino2, struct fs_path *fs_path) { - u64 ino = ino2; - bool free_path = false; + bool free_fs_path = false; int ret = 0; + struct btrfs_path *path = NULL; + struct btrfs_key key; if (!fs_path) { fs_path = fs_path_alloc(); if (!fs_path) return -ENOMEM; - free_path = true; + free_fs_path = true; } - while (ino > BTRFS_FIRST_FREE_OBJECTID) { - u64 parent; - u64 parent_gen; + path = alloc_path_for_send(); + if (!path) { + ret = -ENOMEM; + goto out; + } - fs_path_reset(fs_path); - ret = get_first_ref(root, ino, &parent, &parent_gen, fs_path); - if (ret < 0) { - if (ret == -ENOENT && ino == ino2) - ret = 0; - goto out; + key.objectid = ino2; + key.type = BTRFS_INODE_REF_KEY; + key.offset = 0; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + goto out; + + while (true) { + struct extent_buffer *leaf = path->nodes[0]; + int slot = path->slots[0]; + u32 cur_offset = 0; + u32 item_size; + + if (slot >= btrfs_header_nritems(leaf)) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) + goto out; + if (ret > 0) + break; + continue; } - if (parent == ino1) { - ret = parent_gen == ino1_gen ? 1 : 0; - goto out; + + btrfs_item_key_to_cpu(leaf, &key, slot); + if (key.objectid != ino2) + break; + if (key.type != BTRFS_INODE_REF_KEY && + key.type != BTRFS_INODE_EXTREF_KEY) + break; + + item_size = btrfs_item_size_nr(leaf, slot); + while (cur_offset < item_size) { + u64 parent; + u64 parent_gen; + + if (key.type == BTRFS_INODE_EXTREF_KEY) { + unsigned long ptr; + struct btrfs_inode_extref *extref; + + ptr = btrfs_item_ptr_offset(leaf, slot); + extref = (struct btrfs_inode_extref *) + (ptr + cur_offset); + parent = btrfs_inode_extref_parent(leaf, + extref); + cur_offset += sizeof(*extref); + cur_offset += btrfs_inode_extref_name_len(leaf, + extref); + } else { + parent = key.offset; + cur_offset = item_size; + } + + ret = get_inode_info(root, parent, NULL, &parent_gen, + NULL, NULL, NULL, NULL); + if (ret < 0) + goto out; + ret = check_ino_in_path(root, ino1, ino1_gen, + parent, parent_gen, fs_path); + if (ret) + goto out; } - ino = parent; + path->slots[0]++; } + ret = 0; out: - if (free_path) + btrfs_free_path(path); + if (free_fs_path) fs_path_free(fs_path); return ret; } Patches currently in stable-queue which might be from fdmanana(a)suse.com are queue-4.14/btrfs-incremental-send-fix-wrong-unlink-path-after-renaming-file.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "btrfs: Fix transaction abort during failure in btrfs_rm_dev_item" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled btrfs: Fix transaction abort during failure in btrfs_rm_dev_item to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: btrfs-fix-transaction-abort-during-failure-in-btrfs_rm_dev_item.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: Nikolay Borisov <nborisov(a)suse.com> Date: Mon, 23 Oct 2017 09:58:46 +0300 Subject: btrfs: Fix transaction abort during failure in btrfs_rm_dev_item From: Nikolay Borisov <nborisov(a)suse.com> [ Upstream commit 5e9f2ad5b2904a7e81df6d9a3dbef29478952eac ] btrfs_rm_dev_item calls several function under an active transaction, however it fails to abort it if an error happens. Fix this by adding explicit btrfs_abort_transaction/btrfs_end_transaction calls. Signed-off-by: Nikolay Borisov <nborisov(a)suse.com> Reviewed-by: David Sterba <dsterba(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- fs/btrfs/volumes.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1765,20 +1765,24 @@ static int btrfs_rm_dev_item(struct btrf key.offset = device->devid; ret = btrfs_search_slot(trans, root, &key, path, -1, 1); - if (ret < 0) - goto out; - - if (ret > 0) { - ret = -ENOENT; + if (ret) { + if (ret > 0) + ret = -ENOENT; + btrfs_abort_transaction(trans, ret); + btrfs_end_transaction(trans); goto out; } ret = btrfs_del_item(trans, root, path); - if (ret) - goto out; + if (ret) { + btrfs_abort_transaction(trans, ret); + btrfs_end_transaction(trans); + } + out: btrfs_free_path(path); - btrfs_commit_transaction(trans); + if (!ret) + ret = btrfs_commit_transaction(trans); return ret; } Patches currently in stable-queue which might be from nborisov(a)suse.com are queue-4.14/btrfs-fix-transaction-abort-during-failure-in-btrfs_rm_dev_item.patch

7 years, 5 months

1
0
0 0

[Linux-stable-mirror] Patch "Btrfs: fix list_add corruption and soft lockups in fsync" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled Btrfs: fix list_add corruption and soft lockups in fsync to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: btrfs-fix-list_add-corruption-and-soft-lockups-in-fsync.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Thu Feb 1 13:45:42 CET 2018 From: Liu Bo <bo.li.liu(a)oracle.com> Date: Tue, 21 Nov 2017 14:35:40 -0700 Subject: Btrfs: fix list_add corruption and soft lockups in fsync From: Liu Bo <bo.li.liu(a)oracle.com> [ Upstream commit ebb70442cdd4872260c2415929c456be3562da82 ] Xfstests btrfs/146 revealed this corruption, [ 58.138831] Buffer I/O error on dev dm-0, logical block 2621424, async page read [ 58.151233] BTRFS error (device sdf): bdev /dev/mapper/error-test errs: wr 1, rd 0, flush 0, corrupt 0, gen 0 [ 58.152403] list_add corruption. prev->next should be next (ffff88005e6775d8), but was ffffc9000189be88. (prev=ffffc9000189be88). [ 58.153518] ------------[ cut here ]------------ [ 58.153892] WARNING: CPU: 1 PID: 1287 at lib/list_debug.c:31 __list_add_valid+0x169/0x1f0 ... [ 58.157379] RIP: 0010:__list_add_valid+0x169/0x1f0 ... [ 58.161956] Call Trace: [ 58.162264] btrfs_log_inode_parent+0x5bd/0xfb0 [btrfs] [ 58.163583] btrfs_log_dentry_safe+0x60/0x80 [btrfs] [ 58.164003] btrfs_sync_file+0x4c2/0x6f0 [btrfs] [ 58.164393] vfs_fsync_range+0x5f/0xd0 [ 58.164898] do_fsync+0x5a/0x90 [ 58.165170] SyS_fsync+0x10/0x20 [ 58.165395] entry_SYSCALL_64_fastpath+0x1f/0xbe ... It turns out that we could record btrfs_log_ctx:io_err in log_one_extents when IO fails, but make log_one_extents() return '0' instead of -EIO, so the IO error is not acknowledged by the callers, i.e. btrfs_log_inode_parent(), which would remove btrfs_log_ctx:list from list head 'root->log_ctxs'. Since btrfs_log_ctx is allocated from stack memory, it'd get freed with a object alive on the list. then a future list_add will throw the above warning. This returns the correct error in the above case. Jeff also reported this while testing against his fsync error patch set[1]. [1]: https://www.spinics.net/lists/linux-btrfs/msg65308.html "btrfs list corruption and soft lockups while testing writeback error handling" Fixes: 8407f553268a4611f254 ("Btrfs: fix data corruption after fast fsync and writeback error") Signed-off-by: Liu Bo <bo.li.liu(a)oracle.com> Reviewed-by: David Sterba <dsterba(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> Signed-off-by: Sasha Levin <alexander.levin(a)verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- fs/btrfs/file.c | 5 +++-- fs/btrfs/tree-log.c | 2 +- 2 files changed, 4 insertions(+), 3 deletions(-) --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -2056,6 +2056,8 @@ int btrfs_sync_file(struct file *file, l len = (u64)end - (u64)start + 1; trace_btrfs_sync_file(file, datasync); + btrfs_init_log_ctx(&ctx, inode); + /* * We write the dirty pages in the range and wait until they complete * out of the ->i_mutex. If so, we can flush the dirty pages by @@ -2202,8 +2204,6 @@ int btrfs_sync_file(struct file *file, l } trans->sync = true; - btrfs_init_log_ctx(&ctx, inode); - ret = btrfs_log_dentry_safe(trans, root, dentry, start, end, &ctx); if (ret < 0) { /* Fallthrough and commit/free transaction. */ @@ -2261,6 +2261,7 @@ int btrfs_sync_file(struct file *file, l ret = btrfs_end_transaction(trans); } out: + ASSERT(list_empty(&ctx.list)); err = file_check_and_advance_wb_err(file); if (!ret) ret = err; --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -4100,7 +4100,7 @@ static int log_one_extent(struct btrfs_t if (ordered_io_err) { ctx->io_err = -EIO; - return 0; + return ctx->io_err; } btrfs_init_map_token(&token); Patches currently in stable-queue which might be from bo.li.liu(a)oracle.com are queue-4.14/btrfs-bail-out-gracefully-rather-than-bug_on.patch queue-4.14/btrfs-fix-deadlock-when-writing-out-space-cache.patch queue-4.14/btrfs-fix-list_add-corruption-and-soft-lockups-in-fsync.patch

7 years, 5 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror