Backport gntdev based zero-copy to 4.14 from kernel upstream #42
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This reverts commit 4881b23. New upstreamed implementation should be used instead.
[ Upstream commit 6e3cc2a ] Only gnttab_{alloc|free}_pages are exported as EXPORT_SYMBOL while all the rest are exported as EXPORT_SYMBOL_GPL, thus effectively making it not possible for non-GPL driver modules to use grant table module. Export gnttab_{alloc|free}_pages as EXPORT_SYMBOL_GPL so all the exports are aligned. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Juergen Gross <[email protected]> Signed-off-by: Juergen Gross <[email protected]>
[ Upstream commit 8c3799e ] Make set/clear page private code shared and accessible to other kernel modules which can re-use these instead of open-coding. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit ae4c51a with 4.14 fixes ] Memory {increase|decrease}_reservation and VA mappings update/reset code used in balloon driver can be made common, so other drivers can also re-use the same functionality without open-coding. Create a dedicated file for the shared code and export corresponding symbols for other kernel modules. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit 9bdc730 ] Extend grant table module API to allow allocating buffers that can be used for DMA operations and mapping foreign grant references on top of those. The resulting buffer is similar to the one allocated by the balloon driver in that proper memory reservation is made by ({increase|decrease}_reservation and VA mappings are updated if needed). This is useful for sharing foreign buffers with HW drivers which cannot work with scattered buffers provided by the balloon driver, but require DMAable memory instead. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit 975ef7f with 4.14 fixes ] Fixes for 4.14: - provide a fake bus to miscdevice, so of_dma_configure can complete Allow mappings for DMA backed buffers if grant table module supports such: this extends grant device to not only map buffers made of balloon pages, but also from buffers allocated with dma_alloc_xxx. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit 1d31456 ] This is in preparation for adding support of DMA buffer functionality: make map/unmap related code and structures, used privately by gntdev, ready for dma-buf extension, which will re-use these. Rename corresponding structures as those become non-private to gntdev now. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit 932d656 ] Add UAPI and IOCTLs for dma-buf grant device driver extension: the extension allows userspace processes and kernel modules to use Xen backed dma-buf implementation. With this extension grant references to the pages of an imported dma-buf can be exported for other domain use and grant references coming from a foreign domain can be converted into a local dma-buf for local export. Implement basic initialization and stubs for Xen DMA buffers' support. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit a240d6e with fixes for 4.14 ] Fixes for 4.14: - implement struct dma_buf_ops.map_atomic/unmap_atomic 1. Create a dma-buf from grant references provided by the foreign domain. By default dma-buf is backed by system memory pages, but by providing GNTDEV_DMA_FLAG_XXX flags it can also be created as a DMA write-combine/coherent buffer, e.g. allocated with corresponding dma_alloc_xxx API. Export the resulting buffer as a new dma-buf. 2. Implement waiting for the dma-buf to be released: block until the dma-buf with the file descriptor provided is released. If within the time-out provided the buffer is not released then -ETIMEDOUT error is returned. If the buffer with the file descriptor does not exist or has already been released, then -ENOENT is returned. For valid file descriptors this must not be treated as error. 3. Make gntdev's common code and structures available to dma-buf. [boris: added 'args.fd = -1' to dmabuf_exp_from_refs() to avoid an unnecessary warning about it not being initialized on i386 with gcc 8.1.1] Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit bf8dc55 ] 1. Import a dma-buf with the file descriptor provided and export granted references to the pages of that dma-buf into the array of grant references. 2. Add API to close all references to an imported buffer, so it can be released by the owner. This is only valid for buffers created with IOCTL_GNTDEV_DMABUF_IMP_TO_REFS. Signed-off-by: Oleksandr Andrushchenko <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
[ Upstream commit 2789e83 ] Currently when the allocation of gntdev_dmabuf fails, the error exit path will call dmabuf_imp_free_storage and causes a null pointer dereference on gntdev_dmabuf. Fix this by adding an error exit path that won't free gntdev_dmabuf. Detected by CoverityScan, CID#1472124 ("Dereference after null check") Fixes: bf8dc55 ("xen/gntdev: Implement dma-buf import functionality") Signed-off-by: Colin Ian King <[email protected]> Reviewed-by: Boris Ostrovsky <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
andr2000
requested review from
lorc,
arminn,
klogg,
otyshchenko1,
aanisov and
iartemenko
|
@iartemenko pls merge |
andr2000
pushed a commit
to andr2000/linux
that referenced
this pull request
Nov 27, 2018
Increase kasan instrumented kernel stack size from 32k to 64k. Other architectures seems to get away with just doubling kernel stack size under kasan, but on s390 this appears to be not enough due to bigger frame size. The particular pain point is kasan inlined checks (CONFIG_KASAN_INLINE vs CONFIG_KASAN_OUTLINE). With inlined checks one particular case hitting stack overflow is fs sync on xfs filesystem: #0 [9a0681e8] 704 bytes check_usage at 34b1fc #1 [9a0684a8] 432 bytes check_usage at 34c710 #2 [9a068658] 1048 bytes validate_chain at 35044a #3 [9a068a70] 312 bytes __lock_acquire at 3559fe xen-troops#4 [9a068ba8] 440 bytes lock_acquire at 3576ee xen-troops#5 [9a068d60] 104 bytes _raw_spin_lock at 21b44e0 xen-troops#6 [9a068dc8] 1992 bytes enqueue_entity at 2dbf72 xen-troops#7 [9a069590] 1496 bytes enqueue_task_fair at 2df5f0 xen-troops#8 [9a069b68] 64 bytes ttwu_do_activate at 28f438 xen-troops#9 [9a069ba8] 552 bytes try_to_wake_up at 298c4c xen-troops#10 [9a069dd0] 168 bytes wake_up_worker at 23f97c xen-troops#11 [9a069e78] 200 bytes insert_work at 23fc2e xen-troops#12 [9a069f40] 648 bytes __queue_work at 2487c0 xen-troops#13 [9a06a1c8] 200 bytes __queue_delayed_work at 24db28 xen-troops#14 [9a06a290] 248 bytes mod_delayed_work_on at 24de84 xen-troops#15 [9a06a388] 24 bytes kblockd_mod_delayed_work_on at 153e2a0 xen-troops#16 [9a06a3a0] 288 bytes __blk_mq_delay_run_hw_queue at 158168c xen-troops#17 [9a06a4c0] 192 bytes blk_mq_run_hw_queue at 1581a3c xen-troops#18 [9a06a580] 184 bytes blk_mq_sched_insert_requests at 15a2192 xen-troops#19 [9a06a638] 1024 bytes blk_mq_flush_plug_list at 1590f3a xen-troops#20 [9a06aa38] 704 bytes blk_flush_plug_list at 1555028 xen-troops#21 [9a06acf8] 320 bytes schedule at 219e476 xen-troops#22 [9a06ae38] 760 bytes schedule_timeout at 21b0aac xen-troops#23 [9a06b130] 408 bytes wait_for_common at 21a1706 xen-troops#24 [9a06b2c8] 360 bytes xfs_buf_iowait at fa1540 xen-troops#25 [9a06b430] 256 bytes __xfs_buf_submit at fadae6 xen-troops#26 [9a06b530] 264 bytes xfs_buf_read_map at fae3f6 xen-troops#27 [9a06b638] 656 bytes xfs_trans_read_buf_map at 10ac9a8 xen-troops#28 [9a06b8c8] 304 bytes xfs_btree_kill_root at e72426 xen-troops#29 [9a06b9f8] 288 bytes xfs_btree_lookup_get_block at e7bc5e xen-troops#30 [9a06bb18] 624 bytes xfs_btree_lookup at e7e1a6 xen-troops#31 [9a06bd88] 2664 bytes xfs_alloc_ag_vextent_near at dfa070 xen-troops#32 [9a06c7f0] 144 bytes xfs_alloc_ag_vextent at dff3ca xen-troops#33 [9a06c880] 1128 bytes xfs_alloc_vextent at e05fce xen-troops#34 [9a06cce8] 584 bytes xfs_bmap_btalloc at e58342 xen-troops#35 [9a06cf30] 1336 bytes xfs_bmapi_write at e618de xen-troops#36 [9a06d468] 776 bytes xfs_iomap_write_allocate at ff678e xen-troops#37 [9a06d770] 720 bytes xfs_map_blocks at f82af8 xen-troops#38 [9a06da40] 928 bytes xfs_writepage_map at f83cd6 xen-troops#39 [9a06dde0] 320 bytes xfs_do_writepage at f85872 xen-troops#40 [9a06df20] 1320 bytes write_cache_pages at 73dfe8 xen-troops#41 [9a06e448] 208 bytes xfs_vm_writepages at f7f892 xen-troops#42 [9a06e518] 88 bytes do_writepages at 73fe6a xen-troops#43 [9a06e570] 872 bytes __writeback_single_inode at a20cb6 xen-troops#44 [9a06e8d8] 664 bytes writeback_sb_inodes at a23be2 xen-troops#45 [9a06eb70] 296 bytes __writeback_inodes_wb at a242e0 xen-troops#46 [9a06ec98] 928 bytes wb_writeback at a2500e xen-troops#47 [9a06f038] 848 bytes wb_do_writeback at a260ae xen-troops#48 [9a06f388] 536 bytes wb_workfn at a28228 xen-troops#49 [9a06f5a0] 1088 bytes process_one_work at 24a234 xen-troops#50 [9a06f9e0] 1120 bytes worker_thread at 24ba26 xen-troops#51 [9a06fe40] 104 bytes kthread at 26545a xen-troops#52 [9a06fea8] kernel_thread_starter at 21b6b62 To be able to increase the stack size to 64k reuse LLILL instruction in __switch_to function to load 64k - STACK_FRAME_OVERHEAD - __PT_SIZE (65192) value as unsigned. Reported-by: Benjamin Block <[email protected]> Reviewed-by: Heiko Carstens <[email protected]> Signed-off-by: Vasily Gorbik <[email protected]> Signed-off-by: Martin Schwidefsky <[email protected]>
otyshchenko1
pushed a commit
to otyshchenko1/linux
that referenced
this pull request
Oct 26, 2021
Commit 4dd0d5c ("ice: add lock around Tx timestamp tracker flush") added a lock around the Tx timestamp tracker flow which is used to cleanup any left over SKBs and prepare for device removal. This lock is problematic because it is being held around a call to ice_clear_phy_tstamp. The clear function takes a mutex to send a PHY write command to firmware. This could lead to a deadlock if the mutex actually sleeps, and causes the following warning on a kernel with preemption debugging enabled: [ 715.419426] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:573 [ 715.427900] in_atomic(): 1, irqs_disabled(): 0, non_block: 0, pid: 3100, name: rmmod [ 715.435652] INFO: lockdep is turned off. [ 715.439591] Preemption disabled at: [ 715.439594] [<0000000000000000>] 0x0 [ 715.446678] CPU: 52 PID: 3100 Comm: rmmod Tainted: G W OE 5.15.0-rc4+ xen-troops#42 bdd7ec3018e725f159ca0d372ce8c2c0e784891c [ 715.458058] Hardware name: Intel Corporation S2600STQ/S2600STQ, BIOS SE5C620.86B.02.01.0010.010620200716 01/06/2020 [ 715.468483] Call Trace: [ 715.470940] dump_stack_lvl+0x6a/0x9a [ 715.474613] ___might_sleep.cold+0x224/0x26a [ 715.478895] __mutex_lock+0xb3/0x1440 [ 715.482569] ? stack_depot_save+0x378/0x500 [ 715.486763] ? ice_sq_send_cmd+0x78/0x14c0 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.494979] ? kfree+0xc1/0x520 [ 715.498128] ? mutex_lock_io_nested+0x12a0/0x12a0 [ 715.502837] ? kasan_set_free_info+0x20/0x30 [ 715.507110] ? __kasan_slab_free+0x10b/0x140 [ 715.511385] ? slab_free_freelist_hook+0xc7/0x220 [ 715.516092] ? kfree+0xc1/0x520 [ 715.519235] ? ice_deinit_lag+0x16c/0x220 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.527359] ? ice_remove+0x1cf/0x6a0 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.535133] ? pci_device_remove+0xab/0x1d0 [ 715.539318] ? __device_release_driver+0x35b/0x690 [ 715.544110] ? driver_detach+0x214/0x2f0 [ 715.548035] ? bus_remove_driver+0x11d/0x2f0 [ 715.552309] ? pci_unregister_driver+0x26/0x250 [ 715.556840] ? ice_module_exit+0xc/0x2f [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.564799] ? __do_sys_delete_module.constprop.0+0x2d8/0x4e0 [ 715.570554] ? do_syscall_64+0x3b/0x90 [ 715.574303] ? entry_SYSCALL_64_after_hwframe+0x44/0xae [ 715.579529] ? start_flush_work+0x542/0x8f0 [ 715.583719] ? ice_sq_send_cmd+0x78/0x14c0 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.591923] ice_sq_send_cmd+0x78/0x14c0 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.599960] ? wait_for_completion_io+0x250/0x250 [ 715.604662] ? lock_acquire+0x196/0x200 [ 715.608504] ? do_raw_spin_trylock+0xa5/0x160 [ 715.612864] ice_sbq_rw_reg+0x1e6/0x2f0 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.620813] ? ice_reset+0x130/0x130 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.628497] ? __debug_check_no_obj_freed+0x1e8/0x3c0 [ 715.633550] ? trace_hardirqs_on+0x1c/0x130 [ 715.637748] ice_write_phy_reg_e810+0x70/0xf0 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.646220] ? do_raw_spin_trylock+0xa5/0x160 [ 715.650581] ? ice_ptp_release+0x910/0x910 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.658797] ? ice_ptp_release+0x255/0x910 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.667013] ice_clear_phy_tstamp+0x2c/0x110 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.675403] ice_ptp_release+0x408/0x910 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.683440] ice_remove+0x560/0x6a0 [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.691037] ? _raw_spin_unlock_irqrestore+0x46/0x73 [ 715.696005] pci_device_remove+0xab/0x1d0 [ 715.700018] __device_release_driver+0x35b/0x690 [ 715.704637] driver_detach+0x214/0x2f0 [ 715.708389] bus_remove_driver+0x11d/0x2f0 [ 715.712489] pci_unregister_driver+0x26/0x250 [ 715.716857] ice_module_exit+0xc/0x2f [ice 9a7e1ec00971c89ecd3fe0d4dc7da2b3786a421d] [ 715.724637] __do_sys_delete_module.constprop.0+0x2d8/0x4e0 [ 715.730210] ? free_module+0x6d0/0x6d0 [ 715.733963] ? task_work_run+0xe1/0x170 [ 715.737803] ? exit_to_user_mode_loop+0x17f/0x1d0 [ 715.742509] ? rcu_read_lock_sched_held+0x12/0x80 [ 715.747215] ? trace_hardirqs_on+0x1c/0x130 [ 715.751401] do_syscall_64+0x3b/0x90 [ 715.754981] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 715.760033] RIP: 0033:0x7f4dfe59000b [ 715.763612] Code: 73 01 c3 48 8b 0d 6d 1e 0c 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa b8 b0 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 3d 1e 0c 00 f7 d8 64 89 01 48 [ 715.782357] RSP: 002b:00007ffe8c891708 EFLAGS: 00000206 ORIG_RAX: 00000000000000b0 [ 715.789923] RAX: ffffffffffffffda RBX: 00005558a20468b0 RCX: 00007f4dfe59000b [ 715.797054] RDX: 000000000000000a RSI: 0000000000000800 RDI: 00005558a2046918 [ 715.804189] RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 [ 715.811319] R10: 00007f4dfe603ac0 R11: 0000000000000206 R12: 00007ffe8c891940 [ 715.818455] R13: 00007ffe8c8920a3 R14: 00005558a20462a0 R15: 00005558a20468b0 Notice that this is the only case where we use the lock in this way. In the cleanup kthread and work kthread the lock is only taken around the bit accesses. This was done intentionally to avoid this kind of issue. The way the lock is used, we only protect ordering of bit sets vs bit clears. The Tx writers in the hot path don't need to be protected against the entire kthread loop. The Tx queues threads only need to ensure that they do not re-use an index that is currently in use. The cleanup loop does not need to block all new set bits, since it will re-queue itself if new timestamps are present. Fix the tracker flow so that it uses the same flow as the standard cleanup thread. In addition, ensure the in_use bitmap actually gets cleared properly. This fixes the warning and also avoids the potential deadlock that might have occurred otherwise. Fixes: 4dd0d5c ("ice: add lock around Tx timestamp tracker flush") Signed-off-by: Jacob Keller <[email protected]> Signed-off-by: Tony Nguyen <[email protected]> Signed-off-by: David S. Miller <[email protected]>
otyshchenko1
pushed a commit
to otyshchenko1/linux
that referenced
this pull request
Nov 22, 2021
Commit 719c571 ("net: make napi_disable() symmetric with enable") accidentally introduced a bug sometimes leading to a kernel BUG when bringing an iface up/down under heavy traffic load. Prior to this commit, napi_disable() was polling n->state until none of (NAPIF_STATE_SCHED | NAPIF_STATE_NPSVC) is set and then always flip them. Now there's a possibility to get away with the NAPIF_STATE_SCHE unset as 'continue' drops us to the cmpxchg() call with an uninitialized variable, rather than straight to another round of the state check. Error path looks like: napi_disable(): unsigned long val, new; /* new is uninitialized */ do { val = READ_ONCE(n->state); /* NAPIF_STATE_NPSVC and/or NAPIF_STATE_SCHED is set */ if (val & (NAPIF_STATE_SCHED | NAPIF_STATE_NPSVC)) { /* true */ usleep_range(20, 200); continue; /* go straight to the condition check */ } new = val | <...> } while (cmpxchg(&n->state, val, new) != val); /* state == val, cmpxchg() writes garbage */ napi_enable(): do { val = READ_ONCE(n->state); BUG_ON(!test_bit(NAPI_STATE_SCHED, &val)); /* 50/50 boom */ <...> while the typical BUG splat is like: [ 172.652461] ------------[ cut here ]------------ [ 172.652462] kernel BUG at net/core/dev.c:6937! [ 172.656914] invalid opcode: 0000 [#1] PREEMPT SMP PTI [ 172.661966] CPU: 36 PID: 2829 Comm: xdp_redirect_cp Tainted: G I 5.15.0 xen-troops#42 [ 172.670222] Hardware name: Intel Corporation S2600WFT/S2600WFT, BIOS SE5C620.86B.02.01.0014.082620210524 08/26/2021 [ 172.680646] RIP: 0010:napi_enable+0x5a/0xd0 [ 172.684832] Code: 07 49 81 cc 00 01 00 00 4c 89 e2 48 89 d8 80 e6 fb f0 48 0f b1 55 10 48 39 c3 74 10 48 8b 5d 10 f6 c7 04 75 3d f6 c3 01 75 b4 <0f> 0b 5b 5d 41 5c c3 65 ff 05 b8 e5 61 53 48 c7 c6 c0 f3 34 ad 48 [ 172.703578] RSP: 0018:ffffa3c9497477a8 EFLAGS: 00010246 [ 172.708803] RAX: ffffa3c96615a014 RBX: 0000000000000000 RCX: ffff8a4b575301a0 < snip > [ 172.782403] Call Trace: [ 172.784857] <TASK> [ 172.786963] ice_up_complete+0x6f/0x210 [ice] [ 172.791349] ice_xdp+0x136/0x320 [ice] [ 172.795108] ? ice_change_mtu+0x180/0x180 [ice] [ 172.799648] dev_xdp_install+0x61/0xe0 [ 172.803401] dev_xdp_attach+0x1e0/0x550 [ 172.807240] dev_change_xdp_fd+0x1e6/0x220 [ 172.811338] do_setlink+0xee8/0x1010 [ 172.814917] rtnl_setlink+0xe5/0x170 [ 172.818499] ? bpf_lsm_binder_set_context_mgr+0x10/0x10 [ 172.823732] ? security_capable+0x36/0x50 < snip > Fix this by replacing 'do { } while (cmpxchg())' with an "infinite" for-loop with an explicit break. From v1 [0]: - just use a for-loop to simplify both the fix and the existing code (Eric). [0] https://lore.kernel.org/netdev/[email protected] Fixes: 719c571 ("net: make napi_disable() symmetric with enable") Suggested-by: Eric Dumazet <[email protected]> # for-loop Signed-off-by: Alexander Lobakin <[email protected]> Reviewed-by: Jesse Brandeburg <[email protected]> Reviewed-by: Eric Dumazet <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jakub Kicinski <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This has already been upstreamed, so I don't expect any review here.
The new implementation REPLACES DRM based zero-copy driver which must not be used in new setups