February 2025 - Linux-stable-mirror

[PATCH RESEND v2] Documentation/no_hz: Remove description that states boot CPU cannot be nohz_full

by Krishanth Jagaduri

From: Oleg Nesterov <oleg(a)redhat.com> sched/isolation: Prevent boot crash when the boot CPU is nohz_full [ Upstream commit 5097cbcb38e6e0d2627c9dde1985e91d2c9f880e ] Documentation/timers/no_hz.rst states that the "nohz_full=" mask must not include the boot CPU, which is no longer true after: commit 08ae95f4fd3b ("nohz_full: Allow the boot CPU to be nohz_full"). However after: aae17ebb53cd ("workqueue: Avoid using isolated cpus' timers on queue_delayed_work") the kernel will crash at boot time in this case; housekeeping_any_cpu() returns an invalid CPU number until smp_init() brings the first housekeeping CPU up. Change housekeeping_any_cpu() to check the result of cpumask_any_and() and return smp_processor_id() in this case. This is just the simple and backportable workaround which fixes the symptom, but smp_processor_id() at boot time should be safe at least for type == HK_TYPE_TIMER, this more or less matches the tick_do_timer_boot_cpu logic. There is no worry about cpu_down(); tick_nohz_cpu_down() will not allow to offline tick_do_timer_cpu (the 1st online housekeeping CPU). [ Apply only documentation changes as commit which causes boot crash when boot CPU is nohz_full is not backported to stable kernels - Krishanth ] Fixes: aae17ebb53cd ("workqueue: Avoid using isolated cpus' timers on queue_delayed_work") Reported-by: Chris von Recklinghausen <crecklin(a)redhat.com> Signed-off-by: Oleg Nesterov <oleg(a)redhat.com> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Reviewed-by: Phil Auld <pauld(a)redhat.com> Acked-by: Frederic Weisbecker <frederic(a)kernel.org> Link: https://lore.kernel.org/r/20240411143905.GA19288@redhat.com Closes: https://lore.kernel.org/all/20240402105847.GA24832@redhat.com/ Cc: stable(a)vger.kernel.org # 5.4+ Signed-off-by: Krishanth Jagaduri <Krishanth.Jagaduri(a)sony.com> --- Hi, Before kernel 6.9, Documentation/timers/no_hz.rst states that "nohz_full=" mask must not include the boot CPU, which is no longer true after commit 08ae95f4fd3b ("nohz_full: Allow the boot CPU to be nohz_full"). When trying LTS kernels between 5.4 and 6.6, we noticed we could use boot CPU as nohz_full but the information in the document was misleading. This was fixed upstream by commit 5097cbcb38e6 ("sched/isolation: Prevent boot crash when the boot CPU is nohz_full"). While it fixes the document description, it also fixes issue introduced by another commit aae17ebb53cd ("workqueue: Avoid using isolated cpus' timers on queue_delayed_work"). It is unlikely that upstream commit as a whole will be backported to stable kernels which does not contain the commit that introduced the issue of boot crash when boot CPU is nohz_full. Could we fix only the document portion in stable kernels 5.4+ that mentions boot CPU cannot be nohz_full? --- Changes in v2: - Add original changelog and trailers to commit message. - Add backport note for why only document portion is modified. - Link to v1: https://lore.kernel.org/r/20250205-send-oss-20250129-v1-1-d404921e6d7e@sony… --- Documentation/timers/no_hz.rst | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/Documentation/timers/no_hz.rst b/Documentation/timers/no_hz.rst index 065db217cb04fc252bbf6a05991296e7f1d3a4c5..16bda468423e88090c0dc467ca7a5c7f3fd2bf02 100644 --- a/Documentation/timers/no_hz.rst +++ b/Documentation/timers/no_hz.rst @@ -129,11 +129,8 @@ adaptive-tick CPUs: At least one non-adaptive-tick CPU must remain online to handle timekeeping tasks in order to ensure that system calls like gettimeofday() returns accurate values on adaptive-tick CPUs. (This is not an issue for CONFIG_NO_HZ_IDLE=y because there are no running -user processes to observe slight drifts in clock rate.) Therefore, the -boot CPU is prohibited from entering adaptive-ticks mode. Specifying a -"nohz_full=" mask that includes the boot CPU will result in a boot-time -error message, and the boot CPU will be removed from the mask. Note that -this means that your system must have at least two CPUs in order for +user processes to observe slight drifts in clock rate.) Note that this +means that your system must have at least two CPUs in order for CONFIG_NO_HZ_FULL=y to do anything for you. Finally, adaptive-ticks CPUs must have their RCU callbacks offloaded. --- base-commit: 219d54332a09e8d8741c1e1982f5eae56099de85 change-id: 20250129-send-oss-20250129-3c42dcf463eb Best regards, -- Krishanth Jagaduri <Krishanth.Jagaduri(a)sony.com>

4 months, 2 weeks

3
2
0 0

FAILED: patch "[PATCH] media: uvcvideo: Remove dangling pointers" 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 221cd51efe4565501a3dbf04cc011b537dcce7fb # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025021035-alarm-cautious-b382@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 221cd51efe4565501a3dbf04cc011b537dcce7fb Mon Sep 17 00:00:00 2001 From: Ricardo Ribalda <ribalda(a)chromium.org> Date: Tue, 3 Dec 2024 21:20:10 +0000 Subject: [PATCH] media: uvcvideo: Remove dangling pointers When an async control is written, we copy a pointer to the file handle that started the operation. That pointer will be used when the device is done. Which could be anytime in the future. If the user closes that file descriptor, its structure will be freed, and there will be one dangling pointer per pending async control, that the driver will try to use. Clean all the dangling pointers during release(). To avoid adding a performance penalty in the most common case (no async operation), a counter has been introduced with some logic to make sure that it is properly handled. Cc: stable(a)vger.kernel.org Fixes: e5225c820c05 ("media: uvcvideo: Send a control event when a Control Change interrupt arrives") Reviewed-by: Hans de Goede <hdegoede(a)redhat.com> Signed-off-by: Ricardo Ribalda <ribalda(a)chromium.org> Reviewed-by: Laurent Pinchart <laurent.pinchart(a)ideasonboard.com> Link: https://lore.kernel.org/r/20241203-uvc-fix-async-v6-3-26c867231118@chromium… Signed-off-by: Laurent Pinchart <laurent.pinchart(a)ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei(a)kernel.org> diff --git a/drivers/media/usb/uvc/uvc_ctrl.c b/drivers/media/usb/uvc/uvc_ctrl.c index b05b84887e51..4837d8df9c03 100644 --- a/drivers/media/usb/uvc/uvc_ctrl.c +++ b/drivers/media/usb/uvc/uvc_ctrl.c @@ -1579,6 +1579,40 @@ static void uvc_ctrl_send_slave_event(struct uvc_video_chain *chain, uvc_ctrl_send_event(chain, handle, ctrl, mapping, val, changes); } +static void uvc_ctrl_set_handle(struct uvc_fh *handle, struct uvc_control *ctrl, + struct uvc_fh *new_handle) +{ + lockdep_assert_held(&handle->chain->ctrl_mutex); + + if (new_handle) { + if (ctrl->handle) + dev_warn_ratelimited(&handle->stream->dev->udev->dev, + "UVC non compliance: Setting an async control with a pending operation."); + + if (new_handle == ctrl->handle) + return; + + if (ctrl->handle) { + WARN_ON(!ctrl->handle->pending_async_ctrls); + if (ctrl->handle->pending_async_ctrls) + ctrl->handle->pending_async_ctrls--; + } + + ctrl->handle = new_handle; + handle->pending_async_ctrls++; + return; + } + + /* Cannot clear the handle for a control not owned by us.*/ + if (WARN_ON(ctrl->handle != handle)) + return; + + ctrl->handle = NULL; + if (WARN_ON(!handle->pending_async_ctrls)) + return; + handle->pending_async_ctrls--; +} + void uvc_ctrl_status_event(struct uvc_video_chain *chain, struct uvc_control *ctrl, const u8 *data) { @@ -1589,7 +1623,8 @@ void uvc_ctrl_status_event(struct uvc_video_chain *chain, mutex_lock(&chain->ctrl_mutex); handle = ctrl->handle; - ctrl->handle = NULL; + if (handle) + uvc_ctrl_set_handle(handle, ctrl, NULL); list_for_each_entry(mapping, &ctrl->info.mappings, list) { s32 value = __uvc_ctrl_get_value(mapping, data); @@ -1863,7 +1898,7 @@ static int uvc_ctrl_commit_entity(struct uvc_device *dev, if (!rollback && handle && ctrl->info.flags & UVC_CTRL_FLAG_ASYNCHRONOUS) - ctrl->handle = handle; + uvc_ctrl_set_handle(handle, ctrl, handle); } return 0; @@ -2772,6 +2807,26 @@ int uvc_ctrl_init_device(struct uvc_device *dev) return 0; } +void uvc_ctrl_cleanup_fh(struct uvc_fh *handle) +{ + struct uvc_entity *entity; + + guard(mutex)(&handle->chain->ctrl_mutex); + + if (!handle->pending_async_ctrls) + return; + + list_for_each_entry(entity, &handle->chain->dev->entities, list) { + for (unsigned int i = 0; i < entity->ncontrols; ++i) { + if (entity->controls[i].handle != handle) + continue; + uvc_ctrl_set_handle(handle, &entity->controls[i], NULL); + } + } + + WARN_ON(handle->pending_async_ctrls); +} + /* * Cleanup device controls. */ diff --git a/drivers/media/usb/uvc/uvc_v4l2.c b/drivers/media/usb/uvc/uvc_v4l2.c index dee6feeba274..93c6cdb23881 100644 --- a/drivers/media/usb/uvc/uvc_v4l2.c +++ b/drivers/media/usb/uvc/uvc_v4l2.c @@ -671,6 +671,8 @@ static int uvc_v4l2_release(struct file *file) uvc_dbg(stream->dev, CALLS, "%s\n", __func__); + uvc_ctrl_cleanup_fh(handle); + /* Only free resources if this is a privileged handle. */ if (uvc_has_privileges(handle)) uvc_queue_release(&stream->queue); diff --git a/drivers/media/usb/uvc/uvcvideo.h b/drivers/media/usb/uvc/uvcvideo.h index 965a789ed03e..5690cfd61e23 100644 --- a/drivers/media/usb/uvc/uvcvideo.h +++ b/drivers/media/usb/uvc/uvcvideo.h @@ -338,7 +338,11 @@ struct uvc_video_chain { struct uvc_entity *processing; /* Processing unit */ struct uvc_entity *selector; /* Selector unit */ - struct mutex ctrl_mutex; /* Protects ctrl.info */ + struct mutex ctrl_mutex; /* + * Protects ctrl.info, + * ctrl.handle and + * uvc_fh.pending_async_ctrls + */ struct v4l2_prio_state prio; /* V4L2 priority state */ u32 caps; /* V4L2 chain-wide caps */ @@ -613,6 +617,7 @@ struct uvc_fh { struct uvc_video_chain *chain; struct uvc_streaming *stream; enum uvc_handle_state state; + unsigned int pending_async_ctrls; }; struct uvc_driver { @@ -798,6 +803,8 @@ int uvc_ctrl_is_accessible(struct uvc_video_chain *chain, u32 v4l2_id, int uvc_xu_ctrl_query(struct uvc_video_chain *chain, struct uvc_xu_control_query *xqry); +void uvc_ctrl_cleanup_fh(struct uvc_fh *handle); + /* Utility functions */ struct usb_host_endpoint *uvc_find_endpoint(struct usb_host_interface *alts, u8 epaddr);

4 months, 2 weeks

2
1
0 0

FAILED: patch "[PATCH] media: uvcvideo: Only save async fh if success" 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 d9fecd096f67a4469536e040a8a10bbfb665918b # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025021006-sharpie-patchwork-f168@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From d9fecd096f67a4469536e040a8a10bbfb665918b Mon Sep 17 00:00:00 2001 From: Ricardo Ribalda <ribalda(a)chromium.org> Date: Tue, 3 Dec 2024 21:20:08 +0000 Subject: [PATCH] media: uvcvideo: Only save async fh if success Now we keep a reference to the active fh for any call to uvc_ctrl_set, regardless if it is an actual set or if it is a just a try or if the device refused the operation. We should only keep the file handle if the device actually accepted applying the operation. Cc: stable(a)vger.kernel.org Fixes: e5225c820c05 ("media: uvcvideo: Send a control event when a Control Change interrupt arrives") Suggested-by: Hans de Goede <hdegoede(a)redhat.com> Reviewed-by: Hans de Goede <hdegoede(a)redhat.com> Reviewed-by: Laurent Pinchart <laurent.pinchart(a)ideasonboard.com> Signed-off-by: Ricardo Ribalda <ribalda(a)chromium.org> Link: https://lore.kernel.org/r/20241203-uvc-fix-async-v6-1-26c867231118@chromium… Signed-off-by: Laurent Pinchart <laurent.pinchart(a)ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei(a)kernel.org> diff --git a/drivers/media/usb/uvc/uvc_ctrl.c b/drivers/media/usb/uvc/uvc_ctrl.c index bab9fdac98e6..e0806641a8d0 100644 --- a/drivers/media/usb/uvc/uvc_ctrl.c +++ b/drivers/media/usb/uvc/uvc_ctrl.c @@ -1811,7 +1811,10 @@ int uvc_ctrl_begin(struct uvc_video_chain *chain) } static int uvc_ctrl_commit_entity(struct uvc_device *dev, - struct uvc_entity *entity, int rollback, struct uvc_control **err_ctrl) + struct uvc_fh *handle, + struct uvc_entity *entity, + int rollback, + struct uvc_control **err_ctrl) { struct uvc_control *ctrl; unsigned int i; @@ -1859,6 +1862,10 @@ static int uvc_ctrl_commit_entity(struct uvc_device *dev, *err_ctrl = ctrl; return ret; } + + if (!rollback && handle && + ctrl->info.flags & UVC_CTRL_FLAG_ASYNCHRONOUS) + ctrl->handle = handle; } return 0; @@ -1895,8 +1902,8 @@ int __uvc_ctrl_commit(struct uvc_fh *handle, int rollback, /* Find the control. */ list_for_each_entry(entity, &chain->entities, chain) { - ret = uvc_ctrl_commit_entity(chain->dev, entity, rollback, - &err_ctrl); + ret = uvc_ctrl_commit_entity(chain->dev, handle, entity, + rollback, &err_ctrl); if (ret < 0) { if (ctrls) ctrls->error_idx = @@ -2046,9 +2053,6 @@ int uvc_ctrl_set(struct uvc_fh *handle, mapping->set(mapping, value, uvc_ctrl_data(ctrl, UVC_CTRL_DATA_CURRENT)); - if (ctrl->info.flags & UVC_CTRL_FLAG_ASYNCHRONOUS) - ctrl->handle = handle; - ctrl->dirty = 1; ctrl->modified = 1; return 0; @@ -2377,7 +2381,7 @@ int uvc_ctrl_restore_values(struct uvc_device *dev) ctrl->dirty = 1; } - ret = uvc_ctrl_commit_entity(dev, entity, 0, NULL); + ret = uvc_ctrl_commit_entity(dev, NULL, entity, 0, NULL); if (ret < 0) return ret; }

4 months, 2 weeks

2
1
0 0

[PATCH] mcb: fix a double free bug in chameleon_parse_gdd()

by Haoxiang Li

In chameleon_parse_gdd(), if mcb_device_register() fails, 'mdev' would be released in mcb_device_register() via put_device(). Thus, goto 'err' label and free 'mdev' again causes a double free. Just return if mcb_device_register() fails. Fixes: 3764e82e5150 ("drivers: Introduce MEN Chameleon Bus") Cc: stable(a)vger.kernel.org Signed-off-by: Haoxiang Li <haoxiang_li2024(a)163.com> --- drivers/mcb/mcb-parse.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mcb/mcb-parse.c b/drivers/mcb/mcb-parse.c index 02a680c73979..bf0d7d58c8b0 100644 --- a/drivers/mcb/mcb-parse.c +++ b/drivers/mcb/mcb-parse.c @@ -96,7 +96,7 @@ static int chameleon_parse_gdd(struct mcb_bus *bus, ret = mcb_device_register(bus, mdev); if (ret < 0) - goto err; + return ret; return 0; -- 2.25.1

4 months, 2 weeks

2
1
0 0

[PATCH v2 4/4] drm/xe: Add staging tree for VM binds

by Thomas Hellström

From: Matthew Brost <matthew.brost(a)intel.com> Concurrent VM bind staging and zapping of PTEs from a userptr notifier do not work because the view of PTEs is not stable. VM binds cannot acquire the notifier lock during staging, as memory allocations are required. To resolve this race condition, use a staging tree for VM binds that is committed only under the userptr notifier lock during the final step of the bind. This ensures a consistent view of the PTEs in the userptr notifier. A follow up may only use staging for VM in fault mode as this is the only mode in which the above race exists. v3: - Drop zap PTE change (Thomas) - s/xe_pt_entry/xe_pt_entry_staging (Thomas) Suggested-by: Thomas Hellström <thomas.hellstrom(a)linux.intel.com> Cc: <stable(a)vger.kernel.org> Fixes: e8babb280b5e ("drm/xe: Convert multiple bind ops into single job") Fixes: a708f6501c69 ("drm/xe: Update PT layer with better error handling") Signed-off-by: Matthew Brost <matthew.brost(a)intel.com> --- drivers/gpu/drm/xe/xe_pt.c | 58 +++++++++++++++++++++++---------- drivers/gpu/drm/xe/xe_pt_walk.c | 3 +- drivers/gpu/drm/xe/xe_pt_walk.h | 4 +++ 3 files changed, 46 insertions(+), 19 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_pt.c b/drivers/gpu/drm/xe/xe_pt.c index 12a627a23eb4..dc24baa84092 100644 --- a/drivers/gpu/drm/xe/xe_pt.c +++ b/drivers/gpu/drm/xe/xe_pt.c @@ -28,6 +28,8 @@ struct xe_pt_dir { struct xe_pt pt; /** @children: Array of page-table child nodes */ struct xe_ptw *children[XE_PDES]; + /** @staging: Array of page-table staging nodes */ + struct xe_ptw *staging[XE_PDES]; }; #if IS_ENABLED(CONFIG_DRM_XE_DEBUG_VM) @@ -48,9 +50,10 @@ static struct xe_pt_dir *as_xe_pt_dir(struct xe_pt *pt) return container_of(pt, struct xe_pt_dir, pt); } -static struct xe_pt *xe_pt_entry(struct xe_pt_dir *pt_dir, unsigned int index) +static struct xe_pt * +xe_pt_entry_staging(struct xe_pt_dir *pt_dir, unsigned int index) { - return container_of(pt_dir->children[index], struct xe_pt, base); + return container_of(pt_dir->staging[index], struct xe_pt, base); } static u64 __xe_pt_empty_pte(struct xe_tile *tile, struct xe_vm *vm, @@ -125,6 +128,7 @@ struct xe_pt *xe_pt_create(struct xe_vm *vm, struct xe_tile *tile, } pt->bo = bo; pt->base.children = level ? as_xe_pt_dir(pt)->children : NULL; + pt->base.staging = level ? as_xe_pt_dir(pt)->staging : NULL; if (vm->xef) xe_drm_client_add_bo(vm->xef->client, pt->bo); @@ -206,8 +210,8 @@ void xe_pt_destroy(struct xe_pt *pt, u32 flags, struct llist_head *deferred) struct xe_pt_dir *pt_dir = as_xe_pt_dir(pt); for (i = 0; i < XE_PDES; i++) { - if (xe_pt_entry(pt_dir, i)) - xe_pt_destroy(xe_pt_entry(pt_dir, i), flags, + if (xe_pt_entry_staging(pt_dir, i)) + xe_pt_destroy(xe_pt_entry_staging(pt_dir, i), flags, deferred); } } @@ -376,8 +380,10 @@ xe_pt_insert_entry(struct xe_pt_stage_bind_walk *xe_walk, struct xe_pt *parent, /* Continue building a non-connected subtree. */ struct iosys_map *map = &parent->bo->vmap; - if (unlikely(xe_child)) + if (unlikely(xe_child)) { parent->base.children[offset] = &xe_child->base; + parent->base.staging[offset] = &xe_child->base; + } xe_pt_write(xe_walk->vm->xe, map, offset, pte); parent->num_live++; @@ -614,6 +620,7 @@ xe_pt_stage_bind(struct xe_tile *tile, struct xe_vma *vma, .ops = &xe_pt_stage_bind_ops, .shifts = xe_normal_pt_shifts, .max_level = XE_PT_HIGHEST_LEVEL, + .staging = true, }, .vm = xe_vma_vm(vma), .tile = tile, @@ -873,7 +880,7 @@ static void xe_pt_cancel_bind(struct xe_vma *vma, } } -static void xe_pt_commit_locks_assert(struct xe_vma *vma) +static void xe_pt_commit_prepare_locks_assert(struct xe_vma *vma) { struct xe_vm *vm = xe_vma_vm(vma); @@ -885,6 +892,16 @@ static void xe_pt_commit_locks_assert(struct xe_vma *vma) xe_vm_assert_held(vm); } +static void xe_pt_commit_locks_assert(struct xe_vma *vma) +{ + struct xe_vm *vm = xe_vma_vm(vma); + + xe_pt_commit_prepare_locks_assert(vma); + + if (xe_vma_is_userptr(vma)) + lockdep_assert_held_read(&vm->userptr.notifier_lock); +} + static void xe_pt_commit(struct xe_vma *vma, struct xe_vm_pgtable_update *entries, u32 num_entries, struct llist_head *deferred) @@ -895,13 +912,17 @@ static void xe_pt_commit(struct xe_vma *vma, for (i = 0; i < num_entries; i++) { struct xe_pt *pt = entries[i].pt; + struct xe_pt_dir *pt_dir; if (!pt->level) continue; + pt_dir = as_xe_pt_dir(pt); for (j = 0; j < entries[i].qwords; j++) { struct xe_pt *oldpte = entries[i].pt_entries[j].pt; + int j_ = j + entries[i].ofs; + pt_dir->children[j_] = pt_dir->staging[j_]; xe_pt_destroy(oldpte, xe_vma_vm(vma)->flags, deferred); } } @@ -913,7 +934,7 @@ static void xe_pt_abort_bind(struct xe_vma *vma, { int i, j; - xe_pt_commit_locks_assert(vma); + xe_pt_commit_prepare_locks_assert(vma); for (i = num_entries - 1; i >= 0; --i) { struct xe_pt *pt = entries[i].pt; @@ -928,10 +949,10 @@ static void xe_pt_abort_bind(struct xe_vma *vma, pt_dir = as_xe_pt_dir(pt); for (j = 0; j < entries[i].qwords; j++) { u32 j_ = j + entries[i].ofs; - struct xe_pt *newpte = xe_pt_entry(pt_dir, j_); + struct xe_pt *newpte = xe_pt_entry_staging(pt_dir, j_); struct xe_pt *oldpte = entries[i].pt_entries[j].pt; - pt_dir->children[j_] = oldpte ? &oldpte->base : 0; + pt_dir->staging[j_] = oldpte ? &oldpte->base : 0; xe_pt_destroy(newpte, xe_vma_vm(vma)->flags, NULL); } } @@ -943,7 +964,7 @@ static void xe_pt_commit_prepare_bind(struct xe_vma *vma, { u32 i, j; - xe_pt_commit_locks_assert(vma); + xe_pt_commit_prepare_locks_assert(vma); for (i = 0; i < num_entries; i++) { struct xe_pt *pt = entries[i].pt; @@ -961,10 +982,10 @@ static void xe_pt_commit_prepare_bind(struct xe_vma *vma, struct xe_pt *newpte = entries[i].pt_entries[j].pt; struct xe_pt *oldpte = NULL; - if (xe_pt_entry(pt_dir, j_)) - oldpte = xe_pt_entry(pt_dir, j_); + if (xe_pt_entry_staging(pt_dir, j_)) + oldpte = xe_pt_entry_staging(pt_dir, j_); - pt_dir->children[j_] = &newpte->base; + pt_dir->staging[j_] = &newpte->base; entries[i].pt_entries[j].pt = oldpte; } } @@ -1494,6 +1515,7 @@ static unsigned int xe_pt_stage_unbind(struct xe_tile *tile, struct xe_vma *vma, .ops = &xe_pt_stage_unbind_ops, .shifts = xe_normal_pt_shifts, .max_level = XE_PT_HIGHEST_LEVEL, + .staging = true, }, .tile = tile, .modified_start = xe_vma_start(vma), @@ -1535,7 +1557,7 @@ static void xe_pt_abort_unbind(struct xe_vma *vma, { int i, j; - xe_pt_commit_locks_assert(vma); + xe_pt_commit_prepare_locks_assert(vma); for (i = num_entries - 1; i >= 0; --i) { struct xe_vm_pgtable_update *entry = &entries[i]; @@ -1548,7 +1570,7 @@ static void xe_pt_abort_unbind(struct xe_vma *vma, continue; for (j = entry->ofs; j < entry->ofs + entry->qwords; j++) - pt_dir->children[j] = + pt_dir->staging[j] = entries[i].pt_entries[j - entry->ofs].pt ? &entries[i].pt_entries[j - entry->ofs].pt->base : NULL; } @@ -1561,7 +1583,7 @@ xe_pt_commit_prepare_unbind(struct xe_vma *vma, { int i, j; - xe_pt_commit_locks_assert(vma); + xe_pt_commit_prepare_locks_assert(vma); for (i = 0; i < num_entries; ++i) { struct xe_vm_pgtable_update *entry = &entries[i]; @@ -1575,8 +1597,8 @@ xe_pt_commit_prepare_unbind(struct xe_vma *vma, pt_dir = as_xe_pt_dir(pt); for (j = entry->ofs; j < entry->ofs + entry->qwords; j++) { entry->pt_entries[j - entry->ofs].pt = - xe_pt_entry(pt_dir, j); - pt_dir->children[j] = NULL; + xe_pt_entry_staging(pt_dir, j); + pt_dir->staging[j] = NULL; } } } diff --git a/drivers/gpu/drm/xe/xe_pt_walk.c b/drivers/gpu/drm/xe/xe_pt_walk.c index b8b3d2aea492..be602a763ff3 100644 --- a/drivers/gpu/drm/xe/xe_pt_walk.c +++ b/drivers/gpu/drm/xe/xe_pt_walk.c @@ -74,7 +74,8 @@ int xe_pt_walk_range(struct xe_ptw *parent, unsigned int level, u64 addr, u64 end, struct xe_pt_walk *walk) { pgoff_t offset = xe_pt_offset(addr, level, walk); - struct xe_ptw **entries = parent->children ? parent->children : NULL; + struct xe_ptw **entries = walk->staging ? (parent->staging ?: NULL) : + (parent->children ?: NULL); const struct xe_pt_walk_ops *ops = walk->ops; enum page_walk_action action; struct xe_ptw *child; diff --git a/drivers/gpu/drm/xe/xe_pt_walk.h b/drivers/gpu/drm/xe/xe_pt_walk.h index 5ecc4d2f0f65..5c02c244f7de 100644 --- a/drivers/gpu/drm/xe/xe_pt_walk.h +++ b/drivers/gpu/drm/xe/xe_pt_walk.h @@ -11,12 +11,14 @@ /** * struct xe_ptw - base class for driver pagetable subclassing. * @children: Pointer to an array of children if any. + * @staging: Pointer to an array of staging if any. * * Drivers could subclass this, and if it's a page-directory, typically * embed an array of xe_ptw pointers. */ struct xe_ptw { struct xe_ptw **children; + struct xe_ptw **staging; }; /** @@ -41,6 +43,8 @@ struct xe_pt_walk { * as shared pagetables. */ bool shared_pt_mode; + /** @staging: Walk staging PT structure */ + bool staging; }; /** -- 2.48.1

4 months, 2 weeks

1
0
0 0

[PATCH v2 3/4] drm/xe: Fix fault mode invalidation with unbind

by Thomas Hellström

Fix fault mode invalidation racing with unbind leading to the PTE zapping potentially traversing an invalid page-table tree. Do this by holding the notifier lock across PTE zapping. This might transfer any contention waiting on the notifier seqlock read side to the notifier lock read side, but that shouldn't be a major problem. At the same time get rid of the open-coded invalidation in the bind code by relying on the notifier even when the vma bind is not yet committed. Finally let userptr invalidation call a dedicated xe_vm function performing a full invalidation. Fixes: e8babb280b5e ("drm/xe: Convert multiple bind ops into single job") Cc: Thomas Hellström <thomas.hellstrom(a)linux.intel.com> Cc: Matthew Brost <matthew.brost(a)intel.com> Cc: Matthew Auld <matthew.auld(a)intel.com> Cc: <stable(a)vger.kernel.org> # v6.12+ Signed-off-by: Thomas Hellström <thomas.hellstrom(a)linux.intel.com> --- drivers/gpu/drm/xe/xe_pt.c | 38 ++++---------- drivers/gpu/drm/xe/xe_vm.c | 85 +++++++++++++++++++++----------- drivers/gpu/drm/xe/xe_vm.h | 8 +++ drivers/gpu/drm/xe/xe_vm_types.h | 4 +- 4 files changed, 75 insertions(+), 60 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_pt.c b/drivers/gpu/drm/xe/xe_pt.c index 1ddcc7e79a93..12a627a23eb4 100644 --- a/drivers/gpu/drm/xe/xe_pt.c +++ b/drivers/gpu/drm/xe/xe_pt.c @@ -1213,42 +1213,22 @@ static int vma_check_userptr(struct xe_vm *vm, struct xe_vma *vma, return 0; uvma = to_userptr_vma(vma); - notifier_seq = uvma->userptr.notifier_seq; + if (xe_pt_userptr_inject_eagain(uvma)) + xe_vma_userptr_force_invalidate(uvma); - if (uvma->userptr.initial_bind && !xe_vm_in_fault_mode(vm)) - return 0; + notifier_seq = uvma->userptr.notifier_seq; if (!mmu_interval_read_retry(&uvma->userptr.notifier, - notifier_seq) && - !xe_pt_userptr_inject_eagain(uvma)) + notifier_seq)) return 0; - if (xe_vm_in_fault_mode(vm)) { + if (xe_vm_in_fault_mode(vm)) return -EAGAIN; - } else { - spin_lock(&vm->userptr.invalidated_lock); - list_move_tail(&uvma->userptr.invalidate_link, - &vm->userptr.invalidated); - spin_unlock(&vm->userptr.invalidated_lock); - - if (xe_vm_in_preempt_fence_mode(vm)) { - struct dma_resv_iter cursor; - struct dma_fence *fence; - long err; - - dma_resv_iter_begin(&cursor, xe_vm_resv(vm), - DMA_RESV_USAGE_BOOKKEEP); - dma_resv_for_each_fence_unlocked(&cursor, fence) - dma_fence_enable_sw_signaling(fence); - dma_resv_iter_end(&cursor); - - err = dma_resv_wait_timeout(xe_vm_resv(vm), - DMA_RESV_USAGE_BOOKKEEP, - false, MAX_SCHEDULE_TIMEOUT); - XE_WARN_ON(err <= 0); - } - } + /* + * Just continue the operation since exec or rebind worker + * will take care of rebinding. + */ return 0; } diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c index 6fdc17be619e..dd422ac95dc0 100644 --- a/drivers/gpu/drm/xe/xe_vm.c +++ b/drivers/gpu/drm/xe/xe_vm.c @@ -580,51 +580,26 @@ static void preempt_rebind_work_func(struct work_struct *w) trace_xe_vm_rebind_worker_exit(vm); } -static bool vma_userptr_invalidate(struct mmu_interval_notifier *mni, - const struct mmu_notifier_range *range, - unsigned long cur_seq) +static void __vma_userptr_invalidate(struct xe_vm *vm, struct xe_userptr_vma *uvma) { - struct xe_userptr *userptr = container_of(mni, typeof(*userptr), notifier); - struct xe_userptr_vma *uvma = container_of(userptr, typeof(*uvma), userptr); + struct xe_userptr *userptr = &uvma->userptr; struct xe_vma *vma = &uvma->vma; - struct xe_vm *vm = xe_vma_vm(vma); struct dma_resv_iter cursor; struct dma_fence *fence; long err; - xe_assert(vm->xe, xe_vma_is_userptr(vma)); - trace_xe_vma_userptr_invalidate(vma); - - if (!mmu_notifier_range_blockable(range)) - return false; - - vm_dbg(&xe_vma_vm(vma)->xe->drm, - "NOTIFIER: addr=0x%016llx, range=0x%016llx", - xe_vma_start(vma), xe_vma_size(vma)); - - down_write(&vm->userptr.notifier_lock); - mmu_interval_set_seq(mni, cur_seq); - - /* No need to stop gpu access if the userptr is not yet bound. */ - if (!userptr->initial_bind) { - up_write(&vm->userptr.notifier_lock); - return true; - } - /* * Tell exec and rebind worker they need to repin and rebind this * userptr. */ if (!xe_vm_in_fault_mode(vm) && - !(vma->gpuva.flags & XE_VMA_DESTROYED) && vma->tile_present) { + !(vma->gpuva.flags & XE_VMA_DESTROYED)) { spin_lock(&vm->userptr.invalidated_lock); list_move_tail(&userptr->invalidate_link, &vm->userptr.invalidated); spin_unlock(&vm->userptr.invalidated_lock); } - up_write(&vm->userptr.notifier_lock); - /* * Preempt fences turn into schedule disables, pipeline these. * Note that even in fault mode, we need to wait for binds and @@ -642,11 +617,35 @@ static bool vma_userptr_invalidate(struct mmu_interval_notifier *mni, false, MAX_SCHEDULE_TIMEOUT); XE_WARN_ON(err <= 0); - if (xe_vm_in_fault_mode(vm)) { + if (xe_vm_in_fault_mode(vm) && userptr->initial_bind) { err = xe_vm_invalidate_vma(vma); XE_WARN_ON(err); } +} +static bool vma_userptr_invalidate(struct mmu_interval_notifier *mni, + const struct mmu_notifier_range *range, + unsigned long cur_seq) +{ + struct xe_userptr_vma *uvma = container_of(mni, typeof(*uvma), userptr.notifier); + struct xe_vma *vma = &uvma->vma; + struct xe_vm *vm = xe_vma_vm(vma); + + xe_assert(vm->xe, xe_vma_is_userptr(vma)); + trace_xe_vma_userptr_invalidate(vma); + + if (!mmu_notifier_range_blockable(range)) + return false; + + vm_dbg(&xe_vma_vm(vma)->xe->drm, + "NOTIFIER: addr=0x%016llx, range=0x%016llx", + xe_vma_start(vma), xe_vma_size(vma)); + + down_write(&vm->userptr.notifier_lock); + mmu_interval_set_seq(mni, cur_seq); + + __vma_userptr_invalidate(vm, uvma); + up_write(&vm->userptr.notifier_lock); trace_xe_vma_userptr_invalidate_complete(vma); return true; @@ -656,6 +655,34 @@ static const struct mmu_interval_notifier_ops vma_userptr_notifier_ops = { .invalidate = vma_userptr_invalidate, }; +#if IS_ENABLED(CONFIG_DRM_XE_USERPTR_INVAL_INJECT) +/** + * xe_vma_userptr_force_invalidate() - force invalidate a userptr + * @uvma: The userptr vma to invalidate + * + * Perform a forced userptr invalidation for testing purposes. + */ +void xe_vma_userptr_force_invalidate(struct xe_userptr_vma *uvma) +{ + struct xe_vm *vm = xe_vma_vm(&uvma->vma); + + /* Protect against concurrent userptr pinning */ + lockdep_assert_held(&vm->lock); + /* Protect against concurrent notifiers */ + lockdep_assert_held(&vm->userptr.notifier_lock); + /* + * Protect against concurrent instances of this function and + * the critical exec sections + */ + xe_vm_assert_held(vm); + + if (!mmu_interval_read_retry(&uvma->userptr.notifier, + uvma->userptr.notifier_seq)) + uvma->userptr.notifier_seq -= 2; + __vma_userptr_invalidate(vm, uvma); +} +#endif + int xe_vm_userptr_pin(struct xe_vm *vm) { struct xe_userptr_vma *uvma, *next; diff --git a/drivers/gpu/drm/xe/xe_vm.h b/drivers/gpu/drm/xe/xe_vm.h index 7c8e39049223..f5d835271350 100644 --- a/drivers/gpu/drm/xe/xe_vm.h +++ b/drivers/gpu/drm/xe/xe_vm.h @@ -287,4 +287,12 @@ struct xe_vm_snapshot *xe_vm_snapshot_capture(struct xe_vm *vm); void xe_vm_snapshot_capture_delayed(struct xe_vm_snapshot *snap); void xe_vm_snapshot_print(struct xe_vm_snapshot *snap, struct drm_printer *p); void xe_vm_snapshot_free(struct xe_vm_snapshot *snap); + +#if IS_ENABLED(CONFIG_DRM_XE_USERPTR_INVAL_INJECT) +void xe_vma_userptr_force_invalidate(struct xe_userptr_vma *uvma); +#else +static inline void xe_vma_userptr_force_invalidate(struct xe_userptr_vma *uvma) +{ +} +#endif #endif diff --git a/drivers/gpu/drm/xe/xe_vm_types.h b/drivers/gpu/drm/xe/xe_vm_types.h index 52467b9b5348..1fe79bf23b6b 100644 --- a/drivers/gpu/drm/xe/xe_vm_types.h +++ b/drivers/gpu/drm/xe/xe_vm_types.h @@ -228,8 +228,8 @@ struct xe_vm { * up for revalidation. Protected from access with the * @invalidated_lock. Removing items from the list * additionally requires @lock in write mode, and adding - * items to the list requires the @userptr.notifer_lock in - * write mode. + * items to the list requires either the @userptr.notifer_lock in + * write mode, OR @lock in write mode. */ struct list_head invalidated; } userptr; -- 2.48.1

4 months, 2 weeks

1
0
0 0

[PATCH v2 2/4] drm/xe/vm: Fix a misplaced #endif

by Thomas Hellström

Fix a (harmless) misplaced #endif leading to declarations appearing multiple times. Fixes: 0eb2a18a8fad ("drm/xe: Implement VM snapshot support for BO's and userptr") Cc: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> Cc: José Roberto de Souza <jose.souza(a)intel.com> Cc: <stable(a)vger.kernel.org> # v6.9+ Signed-off-by: Thomas Hellström <thomas.hellstrom(a)linux.intel.com> --- drivers/gpu/drm/xe/xe_vm.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/xe/xe_vm.h b/drivers/gpu/drm/xe/xe_vm.h index f66075f8a6fe..7c8e39049223 100644 --- a/drivers/gpu/drm/xe/xe_vm.h +++ b/drivers/gpu/drm/xe/xe_vm.h @@ -282,9 +282,9 @@ static inline void vm_dbg(const struct drm_device *dev, const char *format, ...) { /* noop */ } #endif -#endif struct xe_vm_snapshot *xe_vm_snapshot_capture(struct xe_vm *vm); void xe_vm_snapshot_capture_delayed(struct xe_vm_snapshot *snap); void xe_vm_snapshot_print(struct xe_vm_snapshot *snap, struct drm_printer *p); void xe_vm_snapshot_free(struct xe_vm_snapshot *snap); +#endif -- 2.48.1

4 months, 2 weeks

1
0
0 0

[PATCH V2] LoongArch: Use polling play_dead() when resuming from hibernation

by Huacai Chen

When CONFIG_RANDOM_KMALLOC_CACHES or other randomization infrastructrue enabled, the idle_task's stack may different between the booting kernel and target kernel. So when resuming from hibernation, an ACTION_BOOT_CPU IPI wakeup the idle instruction in arch_cpu_idle_dead() and jump to the interrupt handler. But since the stack pointer is changed, the interrupt handler cannot restore correct context. So rename the current arch_cpu_idle_dead() to idle_play_dead(), make it as the default version of play_dead(), and the new arch_cpu_idle_dead() call play_dead() directly. For hibernation, implement an arch-specific hibernate_resume_nonboot_cpu_disable() to use the polling version (idle instruction is replace by nop, and irq is disabled) of play_dead(), i.e. poll_play_dead(), to avoid IPI handler corrupting the idle_task's stack when resuming from hibernation. This solution is a little similar to commit 406f992e4a372dafbe3c ("x86 / hibernate: Use hlt_play_dead() when resuming from hibernation"). Cc: stable(a)vger.kernel.org Tested-by: Erpeng Xu <xuerpeng(a)uniontech.com> Tested-by: Yuli Wang <wangyuli(a)uniontech.com> Signed-off-by: Huacai Chen <chenhuacai(a)loongson.cn> --- V2: Fix build for !HIBERNATION and restore to idle_play_dead() if fails. arch/loongarch/kernel/smp.c | 47 ++++++++++++++++++++++++++++++++++++- 1 file changed, 46 insertions(+), 1 deletion(-) diff --git a/arch/loongarch/kernel/smp.c b/arch/loongarch/kernel/smp.c index fbf747447f13..4b24589c0b56 100644 --- a/arch/loongarch/kernel/smp.c +++ b/arch/loongarch/kernel/smp.c @@ -19,6 +19,7 @@ #include <linux/smp.h> #include <linux/threads.h> #include <linux/export.h> +#include <linux/suspend.h> #include <linux/syscore_ops.h> #include <linux/time.h> #include <linux/tracepoint.h> @@ -423,7 +424,7 @@ void loongson_cpu_die(unsigned int cpu) mb(); } -void __noreturn arch_cpu_idle_dead(void) +static void __noreturn idle_play_dead(void) { register uint64_t addr; register void (*init_fn)(void); @@ -447,6 +448,50 @@ void __noreturn arch_cpu_idle_dead(void) BUG(); } +#ifdef CONFIG_HIBERNATION +static void __noreturn poll_play_dead(void) +{ + register uint64_t addr; + register void (*init_fn)(void); + + idle_task_exit(); + __this_cpu_write(cpu_state, CPU_DEAD); + + __smp_mb(); + do { + __asm__ __volatile__("nop\n\t"); + addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); + } while (addr == 0); + + init_fn = (void *)TO_CACHE(addr); + iocsr_write32(0xffffffff, LOONGARCH_IOCSR_IPI_CLEAR); + + init_fn(); + BUG(); +} +#endif + +static void (*play_dead)(void) = idle_play_dead; + +void __noreturn arch_cpu_idle_dead(void) +{ + play_dead(); + BUG(); /* play_dead() doesn't return */ +} + +#ifdef CONFIG_HIBERNATION +int hibernate_resume_nonboot_cpu_disable(void) +{ + int ret; + + play_dead = poll_play_dead; + ret = suspend_disable_secondary_cpus(); + play_dead = idle_play_dead; + + return ret; +} +#endif + #endif /* -- 2.47.1

4 months, 2 weeks

1
0
0 0

[PATCH] arm64: dts: ti: k3-j784s4-j742s2-main-common: Fix serdes_ln_ctrl reg-masks

by Siddharth Vadapalli

Commit under Fixes added the 'idle-states' property for SERDES4 lane muxing without defining the corresponding register offsets and masks for it in the 'mux-reg-masks' property within the 'serdes_ln_ctrl' node. Fix this. Fixes: 7287d423f138 ("arm64: dts: ti: k3-j784s4-main: Add system controller and SERDES lane mux") Cc: stable(a)vger.kernel.org Signed-off-by: Siddharth Vadapalli <s-vadapalli(a)ti.com> --- Hello, This patch is based on commit dd83757f6e68 Merge tag 'bcachefs-2025-02-26' of git://evilpiepirate.org/bcachefs of the master branch of Mainline Linux. Regards, Siddharth. arch/arm64/boot/dts/ti/k3-j784s4-j742s2-main-common.dtsi | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/arm64/boot/dts/ti/k3-j784s4-j742s2-main-common.dtsi b/arch/arm64/boot/dts/ti/k3-j784s4-j742s2-main-common.dtsi index 83bbf94b58d1..a5fefafcba74 100644 --- a/arch/arm64/boot/dts/ti/k3-j784s4-j742s2-main-common.dtsi +++ b/arch/arm64/boot/dts/ti/k3-j784s4-j742s2-main-common.dtsi @@ -84,7 +84,9 @@ serdes_ln_ctrl: mux-controller@4080 { <0x10 0x3>, <0x14 0x3>, /* SERDES1 lane0/1 select */ <0x18 0x3>, <0x1c 0x3>, /* SERDES1 lane2/3 select */ <0x20 0x3>, <0x24 0x3>, /* SERDES2 lane0/1 select */ - <0x28 0x3>, <0x2c 0x3>; /* SERDES2 lane2/3 select */ + <0x28 0x3>, <0x2c 0x3>, /* SERDES2 lane2/3 select */ + <0x30 0x3>, <0x34 0x3>, /* SERDES4 lane0/1 select */ + <0x38 0x3>, <0x3c 0x3>; /* SERDES4 lane2/3 select */ idle-states = <J784S4_SERDES0_LANE0_PCIE1_LANE0>, <J784S4_SERDES0_LANE1_PCIE1_LANE1>, <J784S4_SERDES0_LANE2_IP3_UNUSED>, -- 2.34.1

4 months, 2 weeks

1
1
0 0

[PATCH 6.6 v3 0/3] Set the bpf_net_context before invoking BPF XDP in the TUN driver

by Ricardo Cañuelo Navarro

A private syzbot instance reported "KASAN: slab-use-after-free Read in dev_map_enqueue" under some runtime environments. Upstream patch fecef4cd42c6 ("tun: Assign missing bpf_net_context") fixes the issue. In order to bring this patch to stable v6.6 it's also necessary to bring upstream patch 401cb7dae813 ("net: Reference bpf_redirect_info via task_struct on PREEMPT_RT.") as a dependency. The dependency patch (401cb7dae813 ("net: Reference bpf_redirect_info via task_struct on PREEMPT_RT.")) comes from a patch series [1], the second patch addresses a missing change in the series. Only these two patches were picked up because the purpose of this backport is to fix the particular issue discovered by syzbot. However, maybe Sebastian may consider it's a better idea to backport the whole series instead of only these two patches. I'd also appreciate if you can share your opinion on whether this backport should be applied to other stable branches as well. Both patches needed some manual work in order to be applied on stable, mostly related to changes in the context lines: In the case of 401cb7dae813 ("net: Reference bpf_redirect_info via task_struct on PREEMPT_RT."), the backport addresses the differences in net/core/dev.c:napi_threaded_poll(), busy_poll_stop(), napi_busy_loop() and net_rx_action() between upstream and stable. This allows the patch to be applied without bringing additional dependencies, such as dad6b9770263 ("net: Allow to use SMP threads for backlog NAPI."). The rest of the changes are made to adapt context lines and are unrelated to the purpose of the patch. For fecef4cd42c6 ("tun: Assign missing bpf_net_context"), the backport addresses the changes in function parameters introduced by 7cd1107f48e2a ("bpf, xdp: constify some bpf_prog * function arguments") and 4d2bb0bfe874 ("xdp: rely on skb pointer reference in do_xdp_generic and netif_receive_generic_xdp"). Additionally, upstream commit 9da49aa80d68 ("tun: Add missing bpf_net_ctx_clear() in do_xdp_generic()"), which fixes fecef4cd42c6 ("tun: Assign missing bpf_net_context") is also backported with trivial changes to adapt the differences in the patch context. [1] https://lore.kernel.org/all/20240612170303.3896084-1-bigeasy@linutronix.de/ Signed-off-by: Ricardo Cañuelo Navarro <rcn(a)igalia.com> --- Changes in v3: - Additional patch backported: 9da49aa80d68 ("tun: Add missing bpf_net_ctx_clear() in do_xdp_generic()") which fixes fecef4cd42c6 ("tun: Assign missing bpf_net_context."). Suggested by Sasha's helper bot. - Link to v2: https://lore.kernel.org/r/20250225-20250204-kasan-slab-use-after-free-read-… Changes in v2: - Fix backport for patch 401cb7dae813 ("net: Reference bpf_redirect_info via task_struct on PREEMPT_RT.") in v1. - Add context for the patches and SoB tags. - Extend the recipient list. - Link to v1: https://lore.kernel.org/r/20250224-20250204-kasan-slab-use-after-free-read-… --- Jeongjun Park (1): tun: Add missing bpf_net_ctx_clear() in do_xdp_generic() Sebastian Andrzej Siewior (2): net: Reference bpf_redirect_info via task_struct on PREEMPT_RT. tun: Assign missing bpf_net_context. drivers/net/tun.c | 7 +++++++ include/linux/filter.h | 56 +++++++++++++++++++++++++++++++++++++++++--------- include/linux/sched.h | 3 +++ kernel/bpf/cpumap.c | 3 +++ kernel/bpf/devmap.c | 9 +++++++- kernel/fork.c | 1 + net/bpf/test_run.c | 11 +++++++++- net/core/dev.c | 34 +++++++++++++++++++++++++++++- net/core/filter.c | 44 +++++++++++---------------------------- net/core/lwt_bpf.c | 3 +++ 10 files changed, 126 insertions(+), 45 deletions(-) --- base-commit: c0249d3a0c3cf082d56f4285647ddba19ef604a7 change-id: 20250224-20250204-kasan-slab-use-after-free-read-in-dev_map_enqueue__submit-b907af839805 Cheers, Ricardo

4 months, 2 weeks

3
12
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror February 2025