Re: [PATCH] usb: xhci-mtk: set the dma max_seg_size

From: Robin Murphy
Date: Thu Jun 29 2023 - 15:21:02 EST


On 2023-06-29 19:29, Ricardo Ribalda wrote:
Hi Robin

On Thu, 29 Jun 2023 at 20:11, Robin Murphy <robin.murphy@xxxxxxx> wrote:

On 2023-06-28 22:00, Ricardo Ribalda wrote:
Allow devices to have dma operations beyond 64K, and avoid warnings such
as:

Hang on, is this actually correct? I just had a vague memory of XHCI
having some restrictions, and sure enough according to the spec it
*does* require buffers to be split at 64KB boundaries, since that's the
maximum length a single TRB can encode - that's exactly the kind of
constraint that the max_seg_size abstraction is intended to represent,
so it seems a bit odd to be explicitly claiming a very different value.

Thanks,
Robin.

I think we had a similar discussion for 93915a4170e9 ("xhci-pci: set
the dma max_seg_size")
on
https://lore.kernel.org/all/1fe8f8a7-c88f-0c91-e74f-4d3f2f885c89@xxxxxxxxxxxxxxx/

```
Preferred max segment size of sg list would be 64k as xHC hardware has
64k TRB payload size
limit, but xhci driver will take care of larger segments, splitting
them into 64k chunks.
```

OK, but it still seems off to me to claim to support something that the hardware doesn't support, and the driver has to fake, especially when it's only to paper over a warning which isn't even the driver's fault in the first place.

The aim of the DMA_API_DEBUG_SG warnings wasn't to go round blindly adding dma_set_max_seg_size(UINT_MAX) all over the place, it was always to consider whether the dma_map_sg() call and/or the scatterlist itself are right, just as much as whether the driver may have forgotten to set an appropriate parameter. As I've already raised, in this particular case I think it's actually the debug check that's misplaced, since it's not dma_map_sg() anyway, but as it stands, the implementations of dma_alloc_noncontiguous() are definitely doing the wrong thing with respect to what it is then asking to validate.

Unless there is some known reason to make this change to any USB host controller *other* than that someone sees UVC allocate a >64KB buffer via this path on a system which happens to have that particular HCD, it is not the right change to make.

Thanks,
Robin.



DMA-API: xhci-mtk 11200000.usb: mapping sg segment longer than device claims to support [len=98304] [max=65536]

Signed-off-by: Ricardo Ribalda <ribalda@xxxxxxxxxxxx>
---
Fix warnings such as:

[ 451.089443] ------------[ cut here ]------------
[ 451.089498] DMA-API: xhci-mtk 11200000.usb: mapping sg segment longer than device claims to support [len=98304] [max=65536]
[ 451.089617] WARNING: CPU: 7 PID: 14227 at kernel/dma/debug.c:1163 debug_dma_map_sg+0x5bc/0x950
[ 451.089674] Modules linked in: xfrm_interface tun hci_vhci bridge stp llc veth xt_cgroup xt_MASQUERADE uinput rfcomm ip6table_nat fuse 8021q algif_hash algif_skcipher af_alg r8153_ecm cdc_ether usbnet r8152 mii mtk_vcodec_dec_hw mt7921s mt76_sdio mt7921_common mt76_connac_lib mt76 uvcvideo videobuf2_vmalloc mtk_vcodec_dec v4l2_h264 mtk_vcodec_enc mtk_jpeg v4l2_vp9 videobuf2_dma_contig videobuf2_memops v4l2_mem2mem videobuf2_v4l2 btmtksdio videobuf2_common mtk_vcodec_common btmtk mac80211 snd_sof_mt8186 snd_sof_xtensa_dsp snd_sof_of snd_sof snd_sof_utils mtk_scp mtk_rpmsg rpmsg_core mtk_scp_ipi hid_rmi rmi_core serio bluetooth ecdh_generic ecc cfg80211 lzo_rle lzo_compress zram joydev
[ 451.090285] CPU: 7 PID: 14227 Comm: syz-executor.0 Not tainted 5.15.118-lockdep-19753-g1b0a8b16661d #1 cd3ddfc5e13dbbbea438d3161fcad4d98ec474f4
[ 451.090333] Hardware name: Google Rusty sku196608/196609/196610/196611 board (DT)
[ 451.090356] pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 451.090401] pc : debug_dma_map_sg+0x5bc/0x950
[ 451.090433] lr : debug_dma_map_sg+0x5bc/0x950
[ 451.090462] sp : ffffffc01fdd75e0
[ 451.090479] x29: ffffffc01fdd7640 x28: ffffff80c1280300 x27: 0000000000010000
[ 451.090531] x26: ffffff80c1ec9600 x25: 1ffffff01a749501 x24: ffffff80d3a4a800
[ 451.090581] x23: dfffffc000000000 x22: ffffff80d3a4a80c x21: ffffffc00aae5740
[ 451.090631] x20: ffffffffffffffff x19: ffffff80d3a4a810 x18: 0000000000000000
[ 451.090680] x17: 64206e6168742072 x16: 65676e6f6c20746e x15: 656d676573206773
[ 451.090731] x14: 20676e697070616d x13: 0000000000000001 x12: 0000000000000001
[ 451.090779] x11: 0000000000000000 x10: 0000000000040000 x9 : 3c6fd66e79e32400
[ 451.090828] x8 : 3c6fd66e79e32400 x7 : 0000000000000001 x6 : 0000000000000001
[ 451.090877] x5 : ffffffc01fdd7158 x4 : ffffffc00b64e2a0 x3 : ffffffc008f92adc
[ 451.090926] x2 : 0000000100000000 x1 : ffffff8057afd940 x0 : 000000000000006f
[ 451.090976] Call trace:
[ 451.090994] debug_dma_map_sg+0x5bc/0x950
[ 451.091026] dma_alloc_noncontiguous+0x2f4/0x404
[ 451.091060] uvc_alloc_urb_buffers+0x1e8/0x600 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.091150] uvc_video_start_transfer+0xaf4/0x1628 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.091228] uvc_video_start_streaming+0x154/0x2d8 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.091305] uvc_start_streaming+0x20c/0x3d4 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.091379] vb2_start_streaming+0x118/0x400 [videobuf2_common 252dc8c49960dcb8e329e2787100c89e1899c17f]
[ 451.091446] vb2_core_streamon+0x258/0x360 [videobuf2_common 252dc8c49960dcb8e329e2787100c89e1899c17f]
[ 451.091507] vb2_streamon+0x88/0xbc [videobuf2_v4l2 f4acca89bfe3410cd8f3ca536255fc3877fe63db]
[ 451.091555] uvc_queue_streamon+0x44/0x68 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.091631] uvc_ioctl_streamon+0xd8/0x124 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.091705] v4l_streamon+0x74/0xa8
[ 451.091738] __video_do_ioctl+0x90c/0xa40
[ 451.091769] video_usercopy+0xa44/0x1ef8
[ 451.091799] video_ioctl2+0x44/0x58
[ 451.091830] v4l2_ioctl+0x138/0x164
[ 451.091860] __arm64_sys_ioctl+0x154/0x1d0
[ 451.091892] invoke_syscall+0x98/0x278
[ 451.091923] el0_svc_common+0x214/0x274
[ 451.091953] do_el0_svc+0x9c/0x19c
[ 451.091982] el0_svc+0x5c/0xc0
[ 451.092013] el0t_64_sync_handler+0x78/0x108
[ 451.092045] el0t_64_sync+0x1a4/0x1a8
[ 451.092081] Kernel panic - not syncing: kernel: panic_on_warn set ...
[ 451.092103] CPU: 7 PID: 14227 Comm: syz-executor.0 Not tainted 5.15.118-lockdep-19753-g1b0a8b16661d #1 cd3ddfc5e13dbbbea438d3161fcad4d98ec474f4
[ 451.092148] Hardware name: Google Rusty sku196608/196609/196610/196611 board (DT)
[ 451.092171] Call trace:
[ 451.092186] dump_backtrace+0x0/0x4e8
[ 451.092219] show_stack+0x34/0x44
[ 451.092247] dump_stack_lvl+0xdc/0x11c
[ 451.092278] dump_stack+0x1c/0x48
[ 451.092307] panic+0x2a4/0x7b8
[ 451.092335] check_panic_on_warn+0xb8/0x104
[ 451.092369] __warn+0x16c/0x230
[ 451.092399] report_bug+0x160/0x280
[ 451.092432] bug_handler+0x48/0xb8
[ 451.092466] call_break_hook+0x180/0x1b4
[ 451.092498] brk_handler+0x30/0xbc
[ 451.092529] do_debug_exception+0x16c/0x31c
[ 451.092563] el1_dbg+0x64/0x80
[ 451.092592] el1h_64_sync_handler+0x70/0xb4
[ 451.092624] el1h_64_sync+0x7c/0x80
[ 451.092653] debug_dma_map_sg+0x5bc/0x950
[ 451.092685] dma_alloc_noncontiguous+0x2f4/0x404
[ 451.092717] uvc_alloc_urb_buffers+0x1e8/0x600 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.092794] uvc_video_start_transfer+0xaf4/0x1628 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.092868] uvc_video_start_streaming+0x154/0x2d8 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.092942] uvc_start_streaming+0x20c/0x3d4 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.093015] vb2_start_streaming+0x118/0x400 [videobuf2_common 252dc8c49960dcb8e329e2787100c89e1899c17f]
[ 451.093079] vb2_core_streamon+0x258/0x360 [videobuf2_common 252dc8c49960dcb8e329e2787100c89e1899c17f]
[ 451.093139] vb2_streamon+0x88/0xbc [videobuf2_v4l2 f4acca89bfe3410cd8f3ca536255fc3877fe63db]
[ 451.093187] uvc_queue_streamon+0x44/0x68 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.093261] uvc_ioctl_streamon+0xd8/0x124 [uvcvideo 1a151fdc876854366480a9c6b7aaa4b7999fb493]
[ 451.093334] v4l_streamon+0x74/0xa8
[ 451.093366] __video_do_ioctl+0x90c/0xa40
[ 451.093398] video_usercopy+0xa44/0x1ef8
[ 451.093428] video_ioctl2+0x44/0x58
[ 451.093457] v4l2_ioctl+0x138/0x164
[ 451.093487] __arm64_sys_ioctl+0x154/0x1d0
[ 451.093518] invoke_syscall+0x98/0x278
[ 451.093548] el0_svc_common+0x214/0x274
[ 451.093578] do_el0_svc+0x9c/0x19c
[ 451.093607] el0_svc+0x5c/0xc0
[ 451.093637] el0t_64_sync_handler+0x78/0x108
[ 451.093669] el0t_64_sync+0x1a4/0x1a8
[ 451.093701] SMP: stopping secondary CPUs
[ 451.093777] Kernel Offset: disabled
[ 451.093797] CPU features: 0xc00181c1,a3300e42
[ 451.093822] Memory Limit: none

Signed-off-by: Ricardo Ribalda Delgado <ribalda@xxxxxxxxxxxx>
---
drivers/usb/host/xhci-mtk.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/drivers/usb/host/xhci-mtk.c b/drivers/usb/host/xhci-mtk.c
index 90cf40d6d0c3..605b1e1a5098 100644
--- a/drivers/usb/host/xhci-mtk.c
+++ b/drivers/usb/host/xhci-mtk.c
@@ -643,6 +643,8 @@ static int xhci_mtk_probe(struct platform_device *pdev)
pm_runtime_put_autosuspend(dev);
pm_runtime_forbid(dev);

+ dma_set_max_seg_size(dev, UINT_MAX);
+
return 0;

dealloc_usb3_hcd:

---
base-commit: 1b2c92a1cb2469d8c0079dbf496ab86e22e1cb7c
change-id: 20230628-mtk-usb-bf0059f64bd7

Best regards,