June 2024 - Linux-stable-mirror

FAILED: patch "[PATCH] drm/amd/display: prevent register access while in IPS" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y git checkout FETCH_HEAD git cherry-pick -x 56342da3d8cc15efe9df7f29985ba8d256bdc258 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062453-fretted-sulfur-3c0f@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 56342da3d8cc15efe9df7f29985ba8d256bdc258 Mon Sep 17 00:00:00 2001 From: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Date: Mon, 3 Jun 2024 10:16:45 -0400 Subject: [PATCH] drm/amd/display: prevent register access while in IPS We can't read/write to DCN registers while in IPS. Since, that can cause the system to hang. So, before proceeding with the access in that scenario, force the system out of IPS. Cc: stable(a)vger.kernel.org # 6.6+ Reviewed-by: Roman Li <roman.li(a)amd.com> Signed-off-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index e426adf95d7d..e9ac20bed0f2 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -11437,6 +11437,12 @@ void amdgpu_dm_trigger_timing_sync(struct drm_device *dev) mutex_unlock(&adev->dm.dc_lock); } +static inline void amdgpu_dm_exit_ips_for_hw_access(struct dc *dc) +{ + if (dc->ctx->dmub_srv && !dc->ctx->dmub_srv->idle_exit_counter) + dc_exit_ips_for_hw_access(dc); +} + void dm_write_reg_func(const struct dc_context *ctx, uint32_t address, u32 value, const char *func_name) { @@ -11447,6 +11453,8 @@ void dm_write_reg_func(const struct dc_context *ctx, uint32_t address, return; } #endif + + amdgpu_dm_exit_ips_for_hw_access(ctx->dc); cgs_write_register(ctx->cgs_device, address, value); trace_amdgpu_dc_wreg(&ctx->perf_trace->write_count, address, value); } @@ -11470,6 +11478,8 @@ uint32_t dm_read_reg_func(const struct dc_context *ctx, uint32_t address, return 0; } + amdgpu_dm_exit_ips_for_hw_access(ctx->dc); + value = cgs_read_register(ctx->cgs_device, address); trace_amdgpu_dc_rreg(&ctx->perf_trace->read_count, address, value);

1 year

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: Attempt to avoid empty TUs when endpoint is" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y git checkout FETCH_HEAD git cherry-pick -x c03d770c0b014a3007a5874bf6b3c3e64d32aaac # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062458-impotency-swifter-6e17@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From c03d770c0b014a3007a5874bf6b3c3e64d32aaac Mon Sep 17 00:00:00 2001 From: Michael Strauss <michael.strauss(a)amd.com> Date: Tue, 7 May 2024 12:03:15 -0400 Subject: [PATCH] drm/amd/display: Attempt to avoid empty TUs when endpoint is DPIA [WHY] Empty SST TUs are illegal to transmit over a USB4 DP tunnel. Current policy is to configure stream encoder to pack 2 pixels per pclk even when ODM combine is not in use, allowing seamless dynamic ODM reconfiguration. However, in extreme edge cases where average pixel count per TU is less than 2, this can lead to unexpected empty TU generation during compliance testing. For example, VIC 1 with a 1xHBR3 link configuration will average 1.98 pix/TU. [HOW] Calculate average pixel count per TU, and block 2 pixels per clock if endpoint is a DPIA tunnel and pixel clock is low enough that we will never require 2:1 ODM combine. Cc: stable(a)vger.kernel.org # 6.6+ Reviewed-by: Wenjing Liu <wenjing.liu(a)amd.com> Acked-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Michael Strauss <michael.strauss(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c index 5295f52e4fc8..dcced89c07b3 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c @@ -1439,3 +1439,75 @@ void dcn35_set_long_vblank(struct pipe_ctx **pipe_ctx, } } } + +static bool should_avoid_empty_tu(struct pipe_ctx *pipe_ctx) +{ + /* Calculate average pixel count per TU, return false if under ~2.00 to + * avoid empty TUs. This is only required for DPIA tunneling as empty TUs + * are legal to generate for native DP links. Assume TU size 64 as there + * is currently no scenario where it's reprogrammed from HW default. + * MTPs have no such limitation, so this does not affect MST use cases. + */ + unsigned int pix_clk_mhz; + unsigned int symclk_mhz; + unsigned int avg_pix_per_tu_x1000; + unsigned int tu_size_bytes = 64; + struct dc_crtc_timing *timing = &pipe_ctx->stream->timing; + struct dc_link_settings *link_settings = &pipe_ctx->link_config.dp_link_settings; + const struct dc *dc = pipe_ctx->stream->link->dc; + + if (pipe_ctx->stream->link->ep_type != DISPLAY_ENDPOINT_USB4_DPIA) + return false; + + // Not necessary for MST configurations + if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) + return false; + + pix_clk_mhz = timing->pix_clk_100hz / 10000; + + // If this is true, can't block due to dynamic ODM + if (pix_clk_mhz > dc->clk_mgr->bw_params->clk_table.entries[0].dispclk_mhz) + return false; + + switch (link_settings->link_rate) { + case LINK_RATE_LOW: + symclk_mhz = 162; + break; + case LINK_RATE_HIGH: + symclk_mhz = 270; + break; + case LINK_RATE_HIGH2: + symclk_mhz = 540; + break; + case LINK_RATE_HIGH3: + symclk_mhz = 810; + break; + default: + // We shouldn't be tunneling any other rates, something is wrong + ASSERT(0); + return false; + } + + avg_pix_per_tu_x1000 = (1000 * pix_clk_mhz * tu_size_bytes) + / (symclk_mhz * link_settings->lane_count); + + // Add small empirically-decided margin to account for potential jitter + return (avg_pix_per_tu_x1000 < 2020); +} + +bool dcn35_is_dp_dig_pixel_rate_div_policy(struct pipe_ctx *pipe_ctx) +{ + struct dc *dc = pipe_ctx->stream->ctx->dc; + + if (!is_h_timing_divisible_by_2(pipe_ctx->stream)) + return false; + + if (should_avoid_empty_tu(pipe_ctx)) + return false; + + if (dc_is_dp_signal(pipe_ctx->stream->signal) && !dc->link_srv->dp_is_128b_132b_signal(pipe_ctx) && + dc->debug.enable_dp_dig_pixel_rate_div_policy) + return true; + + return false; +} diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h index a731c8880d60..f0ea7d1511ae 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h @@ -95,4 +95,6 @@ void dcn35_set_static_screen_control(struct pipe_ctx **pipe_ctx, void dcn35_set_long_vblank(struct pipe_ctx **pipe_ctx, int num_pipes, uint32_t v_total_min, uint32_t v_total_max); +bool dcn35_is_dp_dig_pixel_rate_div_policy(struct pipe_ctx *pipe_ctx); + #endif /* __DC_HWSS_DCN35_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c index df3bf77f3fb4..199781233fd5 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c @@ -158,7 +158,7 @@ static const struct hwseq_private_funcs dcn35_private_funcs = { .setup_hpo_hw_control = dcn35_setup_hpo_hw_control, .calculate_dccg_k1_k2_values = dcn32_calculate_dccg_k1_k2_values, .set_pixels_per_cycle = dcn32_set_pixels_per_cycle, - .is_dp_dig_pixel_rate_div_policy = dcn32_is_dp_dig_pixel_rate_div_policy, + .is_dp_dig_pixel_rate_div_policy = dcn35_is_dp_dig_pixel_rate_div_policy, .dsc_pg_control = dcn35_dsc_pg_control, .dsc_pg_status = dcn32_dsc_pg_status, .enable_plane = dcn35_enable_plane,

1 year

1
0
0 0

FAILED: patch "[PATCH] drm/amdgpu: revert "take runtime pm reference when we attach" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y git checkout FETCH_HEAD git cherry-pick -x 8bd82363e2ee2eb3a9a8ea1fa94ebe1900d05a71 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062437-coveted-renounce-4f9a@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8bd82363e2ee2eb3a9a8ea1fa94ebe1900d05a71 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig(a)amd.com> Date: Wed, 5 Jun 2024 13:27:20 +0200 Subject: [PATCH] drm/amdgpu: revert "take runtime pm reference when we attach a buffer" v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit b8c415e3bf98 ("drm/amdgpu: take runtime pm reference when we attach a buffer") and commit 425285d39afd ("drm/amdgpu: add amdgpu runpm usage trace for separate funcs"). Taking a runtime pm reference for DMA-buf is actually completely unnecessary and even dangerous. The problem is that calling pm_runtime_get_sync() from the DMA-buf callbacks is illegal because we have the reservation locked here which is also taken during resume. So this would deadlock. When the buffer is in GTT it is still accessible even when the GPU is powered down and when it is in VRAM the buffer gets migrated to GTT before powering down. The only use case which would make it mandatory to keep the runtime pm reference would be if we pin the buffer into VRAM, and that's not something we currently do. v2: improve the commit message Signed-off-by: Christian König <christian.koenig(a)amd.com> Reviewed-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> CC: stable(a)vger.kernel.org diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c index 055ba2ea4c12..662d0f28f358 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c @@ -41,8 +41,6 @@ #include <linux/dma-buf.h> #include <linux/dma-fence-array.h> #include <linux/pci-p2pdma.h> -#include <linux/pm_runtime.h> -#include "amdgpu_trace.h" /** * amdgpu_dma_buf_attach - &dma_buf_ops.attach implementation @@ -58,42 +56,11 @@ static int amdgpu_dma_buf_attach(struct dma_buf *dmabuf, struct drm_gem_object *obj = dmabuf->priv; struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - int r; if (pci_p2pdma_distance(adev->pdev, attach->dev, false) < 0) attach->peer2peer = false; - r = pm_runtime_get_sync(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(1, __func__); - if (r < 0) - goto out; - return 0; - -out: - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(0, __func__); - return r; -} - -/** - * amdgpu_dma_buf_detach - &dma_buf_ops.detach implementation - * - * @dmabuf: DMA-buf where we remove the attachment from - * @attach: the attachment to remove - * - * Called when an attachment is removed from the DMA-buf. - */ -static void amdgpu_dma_buf_detach(struct dma_buf *dmabuf, - struct dma_buf_attachment *attach) -{ - struct drm_gem_object *obj = dmabuf->priv; - struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); - struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(0, __func__); } /** @@ -267,7 +234,6 @@ static int amdgpu_dma_buf_begin_cpu_access(struct dma_buf *dma_buf, const struct dma_buf_ops amdgpu_dmabuf_ops = { .attach = amdgpu_dma_buf_attach, - .detach = amdgpu_dma_buf_detach, .pin = amdgpu_dma_buf_pin, .unpin = amdgpu_dma_buf_unpin, .map_dma_buf = amdgpu_dma_buf_map, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 10832b470448..bc3ac73b6b8d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -181,7 +181,6 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f, struct amd amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr, seq, flags | AMDGPU_FENCE_FLAG_INT); pm_runtime_get_noresume(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(1, __func__); ptr = &ring->fence_drv.fences[seq & ring->fence_drv.num_fences_mask]; if (unlikely(rcu_dereference_protected(*ptr, 1))) { struct dma_fence *old; @@ -309,7 +308,6 @@ bool amdgpu_fence_process(struct amdgpu_ring *ring) dma_fence_put(fence); pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(0, __func__); } while (last_seq != seq); return true; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index 7aafeb763e5d..383fce40d4dd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -554,21 +554,6 @@ TRACE_EVENT(amdgpu_reset_reg_dumps, __entry->value) ); -TRACE_EVENT(amdgpu_runpm_reference_dumps, - TP_PROTO(uint32_t index, const char *func), - TP_ARGS(index, func), - TP_STRUCT__entry( - __field(uint32_t, index) - __string(func, func) - ), - TP_fast_assign( - __entry->index = index; - __assign_str(func); - ), - TP_printk("amdgpu runpm reference dump 0x%x: 0x%s\n", - __entry->index, - __get_str(func)) -); #undef AMDGPU_JOB_GET_TIMELINE_NAME #endif

1 year

1
0
0 0

FAILED: patch "[PATCH] drm/amdgpu: revert "take runtime pm reference when we attach" failed to apply to 6.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.9.y git checkout FETCH_HEAD git cherry-pick -x 8bd82363e2ee2eb3a9a8ea1fa94ebe1900d05a71 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062432-cinch-refining-262b@gregkh' --subject-prefix 'PATCH 6.9.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 8bd82363e2ee2eb3a9a8ea1fa94ebe1900d05a71 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig(a)amd.com> Date: Wed, 5 Jun 2024 13:27:20 +0200 Subject: [PATCH] drm/amdgpu: revert "take runtime pm reference when we attach a buffer" v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit b8c415e3bf98 ("drm/amdgpu: take runtime pm reference when we attach a buffer") and commit 425285d39afd ("drm/amdgpu: add amdgpu runpm usage trace for separate funcs"). Taking a runtime pm reference for DMA-buf is actually completely unnecessary and even dangerous. The problem is that calling pm_runtime_get_sync() from the DMA-buf callbacks is illegal because we have the reservation locked here which is also taken during resume. So this would deadlock. When the buffer is in GTT it is still accessible even when the GPU is powered down and when it is in VRAM the buffer gets migrated to GTT before powering down. The only use case which would make it mandatory to keep the runtime pm reference would be if we pin the buffer into VRAM, and that's not something we currently do. v2: improve the commit message Signed-off-by: Christian König <christian.koenig(a)amd.com> Reviewed-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> CC: stable(a)vger.kernel.org diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c index 055ba2ea4c12..662d0f28f358 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c @@ -41,8 +41,6 @@ #include <linux/dma-buf.h> #include <linux/dma-fence-array.h> #include <linux/pci-p2pdma.h> -#include <linux/pm_runtime.h> -#include "amdgpu_trace.h" /** * amdgpu_dma_buf_attach - &dma_buf_ops.attach implementation @@ -58,42 +56,11 @@ static int amdgpu_dma_buf_attach(struct dma_buf *dmabuf, struct drm_gem_object *obj = dmabuf->priv; struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - int r; if (pci_p2pdma_distance(adev->pdev, attach->dev, false) < 0) attach->peer2peer = false; - r = pm_runtime_get_sync(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(1, __func__); - if (r < 0) - goto out; - return 0; - -out: - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(0, __func__); - return r; -} - -/** - * amdgpu_dma_buf_detach - &dma_buf_ops.detach implementation - * - * @dmabuf: DMA-buf where we remove the attachment from - * @attach: the attachment to remove - * - * Called when an attachment is removed from the DMA-buf. - */ -static void amdgpu_dma_buf_detach(struct dma_buf *dmabuf, - struct dma_buf_attachment *attach) -{ - struct drm_gem_object *obj = dmabuf->priv; - struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); - struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(0, __func__); } /** @@ -267,7 +234,6 @@ static int amdgpu_dma_buf_begin_cpu_access(struct dma_buf *dma_buf, const struct dma_buf_ops amdgpu_dmabuf_ops = { .attach = amdgpu_dma_buf_attach, - .detach = amdgpu_dma_buf_detach, .pin = amdgpu_dma_buf_pin, .unpin = amdgpu_dma_buf_unpin, .map_dma_buf = amdgpu_dma_buf_map, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 10832b470448..bc3ac73b6b8d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -181,7 +181,6 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f, struct amd amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr, seq, flags | AMDGPU_FENCE_FLAG_INT); pm_runtime_get_noresume(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(1, __func__); ptr = &ring->fence_drv.fences[seq & ring->fence_drv.num_fences_mask]; if (unlikely(rcu_dereference_protected(*ptr, 1))) { struct dma_fence *old; @@ -309,7 +308,6 @@ bool amdgpu_fence_process(struct amdgpu_ring *ring) dma_fence_put(fence); pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - trace_amdgpu_runpm_reference_dumps(0, __func__); } while (last_seq != seq); return true; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index 7aafeb763e5d..383fce40d4dd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -554,21 +554,6 @@ TRACE_EVENT(amdgpu_reset_reg_dumps, __entry->value) ); -TRACE_EVENT(amdgpu_runpm_reference_dumps, - TP_PROTO(uint32_t index, const char *func), - TP_ARGS(index, func), - TP_STRUCT__entry( - __field(uint32_t, index) - __string(func, func) - ), - TP_fast_assign( - __entry->index = index; - __assign_str(func); - ), - TP_printk("amdgpu runpm reference dump 0x%x: 0x%s\n", - __entry->index, - __get_str(func)) -); #undef AMDGPU_JOB_GET_TIMELINE_NAME #endif

1 year

1
0
0 0

[PATCH v3] tracing/net_sched: NULL pointer dereference in perf_trace_qdisc_reset()

by yskelg＠gmail.com

From: Yunseong Kim <yskelg(a)gmail.com> In the TRACE_EVENT(qdisc_reset) NULL dereference occurred from qdisc->dev_queue->dev <NULL> ->name This situation simulated from bunch of veths and Bluetooth dis/reconnection. During qdisc initialization, qdisc was being set to noop_queue. In veth_init_queue, the initial tx_num was reduced back to one, causing the qdisc reset to be called with noop, which led to the kernel panic. I've attached the GitHub gist link that C converted syz-execprogram source code and 3 log of reproduced vmcore-dmesg. https://gist.github.com/yskelg/cc64562873ce249cdd0d5a358b77d740 Yeoreum and I use two fuzzing tool simultaneously. One process with syz-executor : https://github.com/google/syzkaller $ ./syz-execprog -executor=./syz-executor -repeat=1 -sandbox=setuid \ -enable=none -collide=false log1 The other process with perf fuzzer: https://github.com/deater/perf_event_tests/tree/master/fuzzer $ perf_event_tests/fuzzer/perf_fuzzer I think this will happen on the kernel version. Linux kernel version +v6.7.10, +v6.8, +v6.9 and it could happen in v6.10. This occurred from 51270d573a8d. I think this patch is absolutely necessary. Previously, It was showing not intended string value of name. I've reproduced 3 time from my fedora 40 Debug Kernel with any other module or patched. version: 6.10.0-0.rc2.20240608gitdc772f8237f9.29.fc41.aarch64+debug [ 5287.164555] veth0_vlan: left promiscuous mode [ 5287.164929] veth1_macvtap: left promiscuous mode [ 5287.164950] veth0_macvtap: left promiscuous mode [ 5287.164983] veth1_vlan: left promiscuous mode [ 5287.165008] veth0_vlan: left promiscuous mode [ 5287.165450] veth1_macvtap: left promiscuous mode [ 5287.165472] veth0_macvtap: left promiscuous mode [ 5287.165502] veth1_vlan: left promiscuous mode … [ 5297.598240] bridge0: port 2(bridge_slave_1) entered blocking state [ 5297.598262] bridge0: port 2(bridge_slave_1) entered forwarding state [ 5297.598296] bridge0: port 1(bridge_slave_0) entered blocking state [ 5297.598313] bridge0: port 1(bridge_slave_0) entered forwarding state [ 5297.616090] 8021q: adding VLAN 0 to HW filter on device bond0 [ 5297.620405] bridge0: port 1(bridge_slave_0) entered disabled state [ 5297.620730] bridge0: port 2(bridge_slave_1) entered disabled state [ 5297.627247] 8021q: adding VLAN 0 to HW filter on device team0 [ 5297.629636] bridge0: port 1(bridge_slave_0) entered blocking state … [ 5298.002798] bridge_slave_0: left promiscuous mode [ 5298.002869] bridge0: port 1(bridge_slave_0) entered disabled state [ 5298.309444] bond0 (unregistering): (slave bond_slave_0): Releasing backup interface [ 5298.315206] bond0 (unregistering): (slave bond_slave_1): Releasing backup interface [ 5298.320207] bond0 (unregistering): Released all slaves [ 5298.354296] hsr_slave_0: left promiscuous mode [ 5298.360750] hsr_slave_1: left promiscuous mode [ 5298.374889] veth1_macvtap: left promiscuous mode [ 5298.374931] veth0_macvtap: left promiscuous mode [ 5298.374988] veth1_vlan: left promiscuous mode [ 5298.375024] veth0_vlan: left promiscuous mode [ 5299.109741] team0 (unregistering): Port device team_slave_1 removed [ 5299.185870] team0 (unregistering): Port device team_slave_0 removed … [ 5300.155443] Bluetooth: hci3: unexpected cc 0x0c03 length: 249 > 1 [ 5300.155724] Bluetooth: hci3: unexpected cc 0x1003 length: 249 > 9 [ 5300.155988] Bluetooth: hci3: unexpected cc 0x1001 length: 249 > 9 …. [ 5301.075531] team0: Port device team_slave_1 added [ 5301.085515] bridge0: port 1(bridge_slave_0) entered blocking state [ 5301.085531] bridge0: port 1(bridge_slave_0) entered disabled state [ 5301.085588] bridge_slave_0: entered allmulticast mode [ 5301.085800] bridge_slave_0: entered promiscuous mode [ 5301.095617] bridge0: port 1(bridge_slave_0) entered blocking state [ 5301.095633] bridge0: port 1(bridge_slave_0) entered disabled state … [ 5301.149734] bond0: (slave bond_slave_0): Enslaving as an active interface with an up link [ 5301.173234] bond0: (slave bond_slave_0): Enslaving as an active interface with an up link [ 5301.180517] bond0: (slave bond_slave_1): Enslaving as an active interface with an up link [ 5301.193481] hsr_slave_0: entered promiscuous mode [ 5301.204425] hsr_slave_1: entered promiscuous mode [ 5301.210172] debugfs: Directory 'hsr0' with parent 'hsr' already present! [ 5301.210185] Cannot create hsr debugfs directory [ 5301.224061] bond0: (slave bond_slave_1): Enslaving as an active interface with an up link [ 5301.246901] bond0: (slave bond_slave_0): Enslaving as an active interface with an up link [ 5301.255934] team0: Port device team_slave_0 added [ 5301.256480] team0: Port device team_slave_1 added [ 5301.256948] team0: Port device team_slave_0 added … [ 5301.435928] hsr_slave_0: entered promiscuous mode [ 5301.446029] hsr_slave_1: entered promiscuous mode [ 5301.455872] debugfs: Directory 'hsr0' with parent 'hsr' already present! [ 5301.455884] Cannot create hsr debugfs directory [ 5301.502664] hsr_slave_0: entered promiscuous mode [ 5301.513675] hsr_slave_1: entered promiscuous mode [ 5301.526155] debugfs: Directory 'hsr0' with parent 'hsr' already present! [ 5301.526164] Cannot create hsr debugfs directory [ 5301.563662] hsr_slave_0: entered promiscuous mode [ 5301.576129] hsr_slave_1: entered promiscuous mode [ 5301.580259] debugfs: Directory 'hsr0' with parent 'hsr' already present! [ 5301.580270] Cannot create hsr debugfs directory [ 5301.590269] 8021q: adding VLAN 0 to HW filter on device bond0 [ 5301.595872] KASAN: null-ptr-deref in range [0x0000000000000130-0x0000000000000137] [ 5301.595877] Mem abort info: [ 5301.595881] ESR = 0x0000000096000006 [ 5301.595885] EC = 0x25: DABT (current EL), IL = 32 bits [ 5301.595889] SET = 0, FnV = 0 [ 5301.595893] EA = 0, S1PTW = 0 [ 5301.595896] FSC = 0x06: level 2 translation fault [ 5301.595900] Data abort info: [ 5301.595903] ISV = 0, ISS = 0x00000006, ISS2 = 0x00000000 [ 5301.595907] CM = 0, WnR = 0, TnD = 0, TagAccess = 0 [ 5301.595911] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0 [ 5301.595915] [dfff800000000026] address between user and kernel address ranges [ 5301.595971] Internal error: Oops: 0000000096000006 [#1] SMP … [ 5301.596076] CPU: 2 PID: 102769 Comm: syz-executor.3 Kdump: loaded Tainted: G W ------- --- 6.10.0-0.rc2.20240608gitdc772f8237f9.29.fc41.aarch64+debug #1 [ 5301.596080] Hardware name: VMware, Inc. VMware20,1/VBSA, BIOS VMW201.00V.21805430.BA64.2305221830 05/22/2023 [ 5301.596082] pstate: 01400005 (nzcv daif +PAN -UAO -TCO +DIT -SSBS BTYPE=--) [ 5301.596085] pc : strnlen+0x40/0x88 [ 5301.596114] lr : trace_event_get_offsets_qdisc_reset+0x6c/0x2b0 [ 5301.596124] sp : ffff8000beef6b40 [ 5301.596126] x29: ffff8000beef6b40 x28: dfff800000000000 x27: 0000000000000001 [ 5301.596131] x26: 6de1800082c62bd0 x25: 1ffff000110aa9e0 x24: ffff800088554f00 [ 5301.596136] x23: ffff800088554ec0 x22: 0000000000000130 x21: 0000000000000140 [ 5301.596140] x20: dfff800000000000 x19: ffff8000beef6c60 x18: ffff7000115106d8 [ 5301.596143] x17: ffff800121bad000 x16: ffff800080020000 x15: 0000000000000006 [ 5301.596147] x14: 0000000000000002 x13: ffff0001f3ed8d14 x12: ffff700017ddeda5 [ 5301.596151] x11: 1ffff00017ddeda4 x10: ffff700017ddeda4 x9 : ffff800082cc5eec [ 5301.596155] x8 : 0000000000000004 x7 : 00000000f1f1f1f1 x6 : 00000000f2f2f200 [ 5301.596158] x5 : 00000000f3f3f3f3 x4 : ffff700017dded80 x3 : 00000000f204f1f1 [ 5301.596162] x2 : 0000000000000026 x1 : 0000000000000000 x0 : 0000000000000130 [ 5301.596166] Call trace: [ 5301.596175] strnlen+0x40/0x88 [ 5301.596179] trace_event_get_offsets_qdisc_reset+0x6c/0x2b0 [ 5301.596182] perf_trace_qdisc_reset+0xb0/0x538 [ 5301.596184] __traceiter_qdisc_reset+0x68/0xc0 [ 5301.596188] qdisc_reset+0x43c/0x5e8 [ 5301.596190] netif_set_real_num_tx_queues+0x288/0x770 [ 5301.596194] veth_init_queues+0xfc/0x130 [veth] [ 5301.596198] veth_newlink+0x45c/0x850 [veth] [ 5301.596202] rtnl_newlink_create+0x2c8/0x798 [ 5301.596205] __rtnl_newlink+0x92c/0xb60 [ 5301.596208] rtnl_newlink+0xd8/0x130 [ 5301.596211] rtnetlink_rcv_msg+0x2e0/0x890 [ 5301.596214] netlink_rcv_skb+0x1c4/0x380 [ 5301.596225] rtnetlink_rcv+0x20/0x38 [ 5301.596227] netlink_unicast+0x3c8/0x640 [ 5301.596231] netlink_sendmsg+0x658/0xa60 [ 5301.596234] __sock_sendmsg+0xd0/0x180 [ 5301.596243] __sys_sendto+0x1c0/0x280 [ 5301.596246] __arm64_sys_sendto+0xc8/0x150 [ 5301.596249] invoke_syscall+0xdc/0x268 [ 5301.596256] el0_svc_common.constprop.0+0x16c/0x240 [ 5301.596259] do_el0_svc+0x48/0x68 [ 5301.596261] el0_svc+0x50/0x188 [ 5301.596265] el0t_64_sync_handler+0x120/0x130 [ 5301.596268] el0t_64_sync+0x194/0x198 [ 5301.596272] Code: eb15001f 54000120 d343fc02 12000801 (38f46842) [ 5301.596285] SMP: stopping secondary CPUs [ 5301.597053] Starting crashdump kernel... [ 5301.597057] Bye! Yeoreum and I don't know if the patch we wrote will fix the underlying cause, but we think that priority is to prevent kernel panic happening. So, we're sending this patch. Link: https://lore.kernel.org/lkml/20240229143432.273b4871@gandalf.local.home/t/ Fixes: 51270d573a8d ("tracing/net_sched: Fix tracepoints that save qdisc_dev() as a string") Cc: netdev(a)vger.kernel.org Cc: stable(a)vger.kernel.org # +v6.7.10, +v6.8 Signed-off-by: Yunseong Kim <yskelg(a)gmail.com> Signed-off-by: Yeoreum Yun <yeoreum.yun(a)arm.com> --- include/trace/events/qdisc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/trace/events/qdisc.h b/include/trace/events/qdisc.h index f1b5e816e7e5..170b51fbe47a 100644 --- a/include/trace/events/qdisc.h +++ b/include/trace/events/qdisc.h @@ -81,7 +81,7 @@ TRACE_EVENT(qdisc_reset, TP_ARGS(q), TP_STRUCT__entry( - __string( dev, qdisc_dev(q)->name ) + __string(dev, qdisc_dev(q) ? qdisc_dev(q)->name : "noop_queue") __string( kind, q->ops->id ) __field( u32, parent ) __field( u32, handle ) -- 2.45.2

1 year

4
7
0 0

[PATCH] ext2: Verify bitmap and itable block numbers before using them

by Jan Kara

Verify bitmap block numbers and inode table blocks are sane before using them for checking bits in the block bitmap. CC: stable(a)vger.kernel.org Signed-off-by: Jan Kara <jack(a)suse.cz> --- fs/ext2/balloc.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) I plan to merge this patch through my tree. diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c index 1bfd6ab11038..b8cfab8f98b9 100644 --- a/fs/ext2/balloc.c +++ b/fs/ext2/balloc.c @@ -77,26 +77,33 @@ static int ext2_valid_block_bitmap(struct super_block *sb, ext2_grpblk_t next_zero_bit; ext2_fsblk_t bitmap_blk; ext2_fsblk_t group_first_block; + ext2_grpblk_t max_bit; group_first_block = ext2_group_first_block_no(sb, block_group); + max_bit = ext2_group_last_block_no(sb, block_group) - group_first_block; /* check whether block bitmap block number is set */ bitmap_blk = le32_to_cpu(desc->bg_block_bitmap); offset = bitmap_blk - group_first_block; - if (!ext2_test_bit(offset, bh->b_data)) + if (offset < 0 || offset > max_bit || + !ext2_test_bit(offset, bh->b_data)) /* bad block bitmap */ goto err_out; /* check whether the inode bitmap block number is set */ bitmap_blk = le32_to_cpu(desc->bg_inode_bitmap); offset = bitmap_blk - group_first_block; - if (!ext2_test_bit(offset, bh->b_data)) + if (offset < 0 || offset > max_bit || + !ext2_test_bit(offset, bh->b_data)) /* bad block bitmap */ goto err_out; /* check whether the inode table block number is set */ bitmap_blk = le32_to_cpu(desc->bg_inode_table); offset = bitmap_blk - group_first_block; + if (offset < 0 || offset > max_bit || + offset + EXT2_SB(sb)->s_itb_per_group - 1 > max_bit) + goto err_out; next_zero_bit = ext2_find_next_zero_bit(bh->b_data, offset + EXT2_SB(sb)->s_itb_per_group, offset); -- 2.35.3

1 year

1
0
0 0

FAILED: patch "[PATCH] KVM: arm64: Disassociate vcpus from redistributor region on" failed to apply to 5.15-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.15-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y git checkout FETCH_HEAD git cherry-pick -x 0d92e4a7ffd5c42b9fa864692f82476c0bf8bcc8 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062438-dean-sacrifice-9c3b@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 0d92e4a7ffd5c42b9fa864692f82476c0bf8bcc8 Mon Sep 17 00:00:00 2001 From: Marc Zyngier <maz(a)kernel.org> Date: Wed, 5 Jun 2024 18:56:37 +0100 Subject: [PATCH] KVM: arm64: Disassociate vcpus from redistributor region on teardown When tearing down a redistributor region, make sure we don't have any dangling pointer to that region stored in a vcpu. Fixes: e5a35635464b ("kvm: arm64: vgic-v3: Introduce vgic_v3_free_redist_region()") Reported-by: Alexander Potapenko <glider(a)google.com> Reviewed-by: Oliver Upton <oliver.upton(a)linux.dev> Signed-off-by: Marc Zyngier <maz(a)kernel.org> Link: https://lore.kernel.org/r/20240605175637.1635653-1-maz@kernel.org Cc: stable(a)vger.kernel.org diff --git a/arch/arm64/kvm/vgic/vgic-init.c b/arch/arm64/kvm/vgic/vgic-init.c index 8f5b7a3e7009..7f68cf58b978 100644 --- a/arch/arm64/kvm/vgic/vgic-init.c +++ b/arch/arm64/kvm/vgic/vgic-init.c @@ -391,7 +391,7 @@ static void kvm_vgic_dist_destroy(struct kvm *kvm) if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) { list_for_each_entry_safe(rdreg, next, &dist->rd_regions, list) - vgic_v3_free_redist_region(rdreg); + vgic_v3_free_redist_region(kvm, rdreg); INIT_LIST_HEAD(&dist->rd_regions); } else { dist->vgic_cpu_base = VGIC_ADDR_UNDEF; diff --git a/arch/arm64/kvm/vgic/vgic-mmio-v3.c b/arch/arm64/kvm/vgic/vgic-mmio-v3.c index a3983a631b5a..9e50928f5d7d 100644 --- a/arch/arm64/kvm/vgic/vgic-mmio-v3.c +++ b/arch/arm64/kvm/vgic/vgic-mmio-v3.c @@ -919,8 +919,19 @@ static int vgic_v3_alloc_redist_region(struct kvm *kvm, uint32_t index, return ret; } -void vgic_v3_free_redist_region(struct vgic_redist_region *rdreg) +void vgic_v3_free_redist_region(struct kvm *kvm, struct vgic_redist_region *rdreg) { + struct kvm_vcpu *vcpu; + unsigned long c; + + lockdep_assert_held(&kvm->arch.config_lock); + + /* Garbage collect the region */ + kvm_for_each_vcpu(c, vcpu, kvm) { + if (vcpu->arch.vgic_cpu.rdreg == rdreg) + vcpu->arch.vgic_cpu.rdreg = NULL; + } + list_del(&rdreg->list); kfree(rdreg); } @@ -945,7 +956,7 @@ int vgic_v3_set_redist_base(struct kvm *kvm, u32 index, u64 addr, u32 count) mutex_lock(&kvm->arch.config_lock); rdreg = vgic_v3_rdist_region_from_index(kvm, index); - vgic_v3_free_redist_region(rdreg); + vgic_v3_free_redist_region(kvm, rdreg); mutex_unlock(&kvm->arch.config_lock); return ret; } diff --git a/arch/arm64/kvm/vgic/vgic.h b/arch/arm64/kvm/vgic/vgic.h index 6106ebd5ba42..03d356a12377 100644 --- a/arch/arm64/kvm/vgic/vgic.h +++ b/arch/arm64/kvm/vgic/vgic.h @@ -316,7 +316,7 @@ vgic_v3_rd_region_size(struct kvm *kvm, struct vgic_redist_region *rdreg) struct vgic_redist_region *vgic_v3_rdist_region_from_index(struct kvm *kvm, u32 index); -void vgic_v3_free_redist_region(struct vgic_redist_region *rdreg); +void vgic_v3_free_redist_region(struct kvm *kvm, struct vgic_redist_region *rdreg); bool vgic_v3_rdist_overlap(struct kvm *kvm, gpa_t base, size_t size);

1 year

1
0
0 0

FAILED: patch "[PATCH] KVM: Fix a data race on last_boosted_vcpu in" failed to apply to 4.19-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.19-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-4.19.y git checkout FETCH_HEAD git cherry-pick -x 49f683b41f28918df3e51ddc0d928cb2e934ccdb # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062422-composer-engaging-202c@gregkh' --subject-prefix 'PATCH 4.19.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 49f683b41f28918df3e51ddc0d928cb2e934ccdb Mon Sep 17 00:00:00 2001 From: Breno Leitao <leitao(a)debian.org> Date: Fri, 10 May 2024 02:23:52 -0700 Subject: [PATCH] KVM: Fix a data race on last_boosted_vcpu in kvm_vcpu_on_spin() Use {READ,WRITE}_ONCE() to access kvm->last_boosted_vcpu to ensure the loads and stores are atomic. In the extremely unlikely scenario the compiler tears the stores, it's theoretically possible for KVM to attempt to get a vCPU using an out-of-bounds index, e.g. if the write is split into multiple 8-bit stores, and is paired with a 32-bit load on a VM with 257 vCPUs: CPU0 CPU1 last_boosted_vcpu = 0xff; (last_boosted_vcpu = 0x100) last_boosted_vcpu[15:8] = 0x01; i = (last_boosted_vcpu = 0x1ff) last_boosted_vcpu[7:0] = 0x00; vcpu = kvm->vcpu_array[0x1ff]; As detected by KCSAN: BUG: KCSAN: data-race in kvm_vcpu_on_spin [kvm] / kvm_vcpu_on_spin [kvm] write to 0xffffc90025a92344 of 4 bytes by task 4340 on cpu 16: kvm_vcpu_on_spin (arch/x86/kvm/../../../virt/kvm/kvm_main.c:4112) kvm handle_pause (arch/x86/kvm/vmx/vmx.c:5929) kvm_intel vmx_handle_exit (arch/x86/kvm/vmx/vmx.c:? arch/x86/kvm/vmx/vmx.c:6606) kvm_intel vcpu_run (arch/x86/kvm/x86.c:11107 arch/x86/kvm/x86.c:11211) kvm kvm_arch_vcpu_ioctl_run (arch/x86/kvm/x86.c:?) kvm kvm_vcpu_ioctl (arch/x86/kvm/../../../virt/kvm/kvm_main.c:?) kvm __se_sys_ioctl (fs/ioctl.c:52 fs/ioctl.c:904 fs/ioctl.c:890) __x64_sys_ioctl (fs/ioctl.c:890) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) read to 0xffffc90025a92344 of 4 bytes by task 4342 on cpu 4: kvm_vcpu_on_spin (arch/x86/kvm/../../../virt/kvm/kvm_main.c:4069) kvm handle_pause (arch/x86/kvm/vmx/vmx.c:5929) kvm_intel vmx_handle_exit (arch/x86/kvm/vmx/vmx.c:? arch/x86/kvm/vmx/vmx.c:6606) kvm_intel vcpu_run (arch/x86/kvm/x86.c:11107 arch/x86/kvm/x86.c:11211) kvm kvm_arch_vcpu_ioctl_run (arch/x86/kvm/x86.c:?) kvm kvm_vcpu_ioctl (arch/x86/kvm/../../../virt/kvm/kvm_main.c:?) kvm __se_sys_ioctl (fs/ioctl.c:52 fs/ioctl.c:904 fs/ioctl.c:890) __x64_sys_ioctl (fs/ioctl.c:890) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) value changed: 0x00000012 -> 0x00000000 Fixes: 217ece6129f2 ("KVM: use yield_to instead of sleep in kvm_vcpu_on_spin") Cc: stable(a)vger.kernel.org Signed-off-by: Breno Leitao <leitao(a)debian.org> Link: https://lore.kernel.org/r/20240510092353.2261824-1-leitao@debian.org Signed-off-by: Sean Christopherson <seanjc(a)google.com> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 14841acb8b95..843aa68cbcd0 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -4025,12 +4025,13 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me, bool yield_to_kernel_mode) { struct kvm *kvm = me->kvm; struct kvm_vcpu *vcpu; - int last_boosted_vcpu = me->kvm->last_boosted_vcpu; + int last_boosted_vcpu; unsigned long i; int yielded = 0; int try = 3; int pass; + last_boosted_vcpu = READ_ONCE(kvm->last_boosted_vcpu); kvm_vcpu_set_in_spin_loop(me, true); /* * We boost the priority of a VCPU that is runnable but not @@ -4068,7 +4069,7 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me, bool yield_to_kernel_mode) yielded = kvm_vcpu_yield_to(vcpu); if (yielded > 0) { - kvm->last_boosted_vcpu = i; + WRITE_ONCE(kvm->last_boosted_vcpu, i); break; } else if (yielded < 0) { try--;

1 year

1
0
0 0

FAILED: patch "[PATCH] KVM: Fix a data race on last_boosted_vcpu in" failed to apply to 5.4-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.4.y git checkout FETCH_HEAD git cherry-pick -x 49f683b41f28918df3e51ddc0d928cb2e934ccdb # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062420-pendant-avenging-0263@gregkh' --subject-prefix 'PATCH 5.4.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 49f683b41f28918df3e51ddc0d928cb2e934ccdb Mon Sep 17 00:00:00 2001 From: Breno Leitao <leitao(a)debian.org> Date: Fri, 10 May 2024 02:23:52 -0700 Subject: [PATCH] KVM: Fix a data race on last_boosted_vcpu in kvm_vcpu_on_spin() Use {READ,WRITE}_ONCE() to access kvm->last_boosted_vcpu to ensure the loads and stores are atomic. In the extremely unlikely scenario the compiler tears the stores, it's theoretically possible for KVM to attempt to get a vCPU using an out-of-bounds index, e.g. if the write is split into multiple 8-bit stores, and is paired with a 32-bit load on a VM with 257 vCPUs: CPU0 CPU1 last_boosted_vcpu = 0xff; (last_boosted_vcpu = 0x100) last_boosted_vcpu[15:8] = 0x01; i = (last_boosted_vcpu = 0x1ff) last_boosted_vcpu[7:0] = 0x00; vcpu = kvm->vcpu_array[0x1ff]; As detected by KCSAN: BUG: KCSAN: data-race in kvm_vcpu_on_spin [kvm] / kvm_vcpu_on_spin [kvm] write to 0xffffc90025a92344 of 4 bytes by task 4340 on cpu 16: kvm_vcpu_on_spin (arch/x86/kvm/../../../virt/kvm/kvm_main.c:4112) kvm handle_pause (arch/x86/kvm/vmx/vmx.c:5929) kvm_intel vmx_handle_exit (arch/x86/kvm/vmx/vmx.c:? arch/x86/kvm/vmx/vmx.c:6606) kvm_intel vcpu_run (arch/x86/kvm/x86.c:11107 arch/x86/kvm/x86.c:11211) kvm kvm_arch_vcpu_ioctl_run (arch/x86/kvm/x86.c:?) kvm kvm_vcpu_ioctl (arch/x86/kvm/../../../virt/kvm/kvm_main.c:?) kvm __se_sys_ioctl (fs/ioctl.c:52 fs/ioctl.c:904 fs/ioctl.c:890) __x64_sys_ioctl (fs/ioctl.c:890) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) read to 0xffffc90025a92344 of 4 bytes by task 4342 on cpu 4: kvm_vcpu_on_spin (arch/x86/kvm/../../../virt/kvm/kvm_main.c:4069) kvm handle_pause (arch/x86/kvm/vmx/vmx.c:5929) kvm_intel vmx_handle_exit (arch/x86/kvm/vmx/vmx.c:? arch/x86/kvm/vmx/vmx.c:6606) kvm_intel vcpu_run (arch/x86/kvm/x86.c:11107 arch/x86/kvm/x86.c:11211) kvm kvm_arch_vcpu_ioctl_run (arch/x86/kvm/x86.c:?) kvm kvm_vcpu_ioctl (arch/x86/kvm/../../../virt/kvm/kvm_main.c:?) kvm __se_sys_ioctl (fs/ioctl.c:52 fs/ioctl.c:904 fs/ioctl.c:890) __x64_sys_ioctl (fs/ioctl.c:890) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) value changed: 0x00000012 -> 0x00000000 Fixes: 217ece6129f2 ("KVM: use yield_to instead of sleep in kvm_vcpu_on_spin") Cc: stable(a)vger.kernel.org Signed-off-by: Breno Leitao <leitao(a)debian.org> Link: https://lore.kernel.org/r/20240510092353.2261824-1-leitao@debian.org Signed-off-by: Sean Christopherson <seanjc(a)google.com> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 14841acb8b95..843aa68cbcd0 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -4025,12 +4025,13 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me, bool yield_to_kernel_mode) { struct kvm *kvm = me->kvm; struct kvm_vcpu *vcpu; - int last_boosted_vcpu = me->kvm->last_boosted_vcpu; + int last_boosted_vcpu; unsigned long i; int yielded = 0; int try = 3; int pass; + last_boosted_vcpu = READ_ONCE(kvm->last_boosted_vcpu); kvm_vcpu_set_in_spin_loop(me, true); /* * We boost the priority of a VCPU that is runnable but not @@ -4068,7 +4069,7 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me, bool yield_to_kernel_mode) yielded = kvm_vcpu_yield_to(vcpu); if (yielded > 0) { - kvm->last_boosted_vcpu = i; + WRITE_ONCE(kvm->last_boosted_vcpu, i); break; } else if (yielded < 0) { try--;

1 year

1
0
0 0

FAILED: patch "[PATCH] KVM: Fix a data race on last_boosted_vcpu in" failed to apply to 5.10-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.10-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.10.y git checkout FETCH_HEAD git cherry-pick -x 49f683b41f28918df3e51ddc0d928cb2e934ccdb # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024062419-goes-tingling-5681@gregkh' --subject-prefix 'PATCH 5.10.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 49f683b41f28918df3e51ddc0d928cb2e934ccdb Mon Sep 17 00:00:00 2001 From: Breno Leitao <leitao(a)debian.org> Date: Fri, 10 May 2024 02:23:52 -0700 Subject: [PATCH] KVM: Fix a data race on last_boosted_vcpu in kvm_vcpu_on_spin() Use {READ,WRITE}_ONCE() to access kvm->last_boosted_vcpu to ensure the loads and stores are atomic. In the extremely unlikely scenario the compiler tears the stores, it's theoretically possible for KVM to attempt to get a vCPU using an out-of-bounds index, e.g. if the write is split into multiple 8-bit stores, and is paired with a 32-bit load on a VM with 257 vCPUs: CPU0 CPU1 last_boosted_vcpu = 0xff; (last_boosted_vcpu = 0x100) last_boosted_vcpu[15:8] = 0x01; i = (last_boosted_vcpu = 0x1ff) last_boosted_vcpu[7:0] = 0x00; vcpu = kvm->vcpu_array[0x1ff]; As detected by KCSAN: BUG: KCSAN: data-race in kvm_vcpu_on_spin [kvm] / kvm_vcpu_on_spin [kvm] write to 0xffffc90025a92344 of 4 bytes by task 4340 on cpu 16: kvm_vcpu_on_spin (arch/x86/kvm/../../../virt/kvm/kvm_main.c:4112) kvm handle_pause (arch/x86/kvm/vmx/vmx.c:5929) kvm_intel vmx_handle_exit (arch/x86/kvm/vmx/vmx.c:? arch/x86/kvm/vmx/vmx.c:6606) kvm_intel vcpu_run (arch/x86/kvm/x86.c:11107 arch/x86/kvm/x86.c:11211) kvm kvm_arch_vcpu_ioctl_run (arch/x86/kvm/x86.c:?) kvm kvm_vcpu_ioctl (arch/x86/kvm/../../../virt/kvm/kvm_main.c:?) kvm __se_sys_ioctl (fs/ioctl.c:52 fs/ioctl.c:904 fs/ioctl.c:890) __x64_sys_ioctl (fs/ioctl.c:890) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) read to 0xffffc90025a92344 of 4 bytes by task 4342 on cpu 4: kvm_vcpu_on_spin (arch/x86/kvm/../../../virt/kvm/kvm_main.c:4069) kvm handle_pause (arch/x86/kvm/vmx/vmx.c:5929) kvm_intel vmx_handle_exit (arch/x86/kvm/vmx/vmx.c:? arch/x86/kvm/vmx/vmx.c:6606) kvm_intel vcpu_run (arch/x86/kvm/x86.c:11107 arch/x86/kvm/x86.c:11211) kvm kvm_arch_vcpu_ioctl_run (arch/x86/kvm/x86.c:?) kvm kvm_vcpu_ioctl (arch/x86/kvm/../../../virt/kvm/kvm_main.c:?) kvm __se_sys_ioctl (fs/ioctl.c:52 fs/ioctl.c:904 fs/ioctl.c:890) __x64_sys_ioctl (fs/ioctl.c:890) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) value changed: 0x00000012 -> 0x00000000 Fixes: 217ece6129f2 ("KVM: use yield_to instead of sleep in kvm_vcpu_on_spin") Cc: stable(a)vger.kernel.org Signed-off-by: Breno Leitao <leitao(a)debian.org> Link: https://lore.kernel.org/r/20240510092353.2261824-1-leitao@debian.org Signed-off-by: Sean Christopherson <seanjc(a)google.com> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 14841acb8b95..843aa68cbcd0 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -4025,12 +4025,13 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me, bool yield_to_kernel_mode) { struct kvm *kvm = me->kvm; struct kvm_vcpu *vcpu; - int last_boosted_vcpu = me->kvm->last_boosted_vcpu; + int last_boosted_vcpu; unsigned long i; int yielded = 0; int try = 3; int pass; + last_boosted_vcpu = READ_ONCE(kvm->last_boosted_vcpu); kvm_vcpu_set_in_spin_loop(me, true); /* * We boost the priority of a VCPU that is runnable but not @@ -4068,7 +4069,7 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me, bool yield_to_kernel_mode) yielded = kvm_vcpu_yield_to(vcpu); if (yielded > 0) { - kvm->last_boosted_vcpu = i; + WRITE_ONCE(kvm->last_boosted_vcpu, i); break; } else if (yielded < 0) { try--;

1 year

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror June 2024