* hfi1 broken due to dma_device changes in ib_register
@ 2020-10-29 21:54 Dennis Dalessandro
2020-10-29 22:01 ` Jason Gunthorpe
0 siblings, 1 reply; 4+ messages in thread
From: Dennis Dalessandro @ 2020-10-29 21:54 UTC (permalink / raw)
To: linux-rdma, Jason Gunthorpe
Just a heads up, 5.10-rc1 is broken for rdmavt/hfi1 after:
e0477b34d9d11 "(RDMA: Explicitly pass in the dma_device to
ib_register_device)"
Running with that change causes the call trace below. Reverting the
patch works around the problem. I haven't yet had a chance to look at
what the actual cause is, but will and follow up with a proposed patch
hopefully soon.
[ 61.331005] ------------[ cut here ]------------
[ 61.336590] WARNING: CPU: 0 PID: 155 at kernel/dma/mapping.c:149
dma_map_page_attrs+0x145/0x1d0
[ 61.346735] Modules linked in: rpcrdma ib_isert iscsi_target_mod
target_core_mod ib_iser libiscsi scsi_transport_iscsi ib_ipoib rdma_ucm
ib_umad rdma_cm hfi1(+) ib_cm iw_cm rdmavt ib_uverbs ib_core
rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfs lockd grace nfs_ssc
fscache sunrpc dm_mirror dm_region_hash dm_log dm_mod iTCO_wdt
iTCO_vendor_support mxm_wmi sb_edac x86_pkg_temp_thermal
intel_powerclamp coretemp crct10dif_pclmul mgag200 crc32_pclmul
i2c_algo_bit ghash_clmulni_intel drm_kms_helper ipmi_si aesni_intel
syscopyarea crypto_simd sysfillrect sysimgblt ipmi_devintf cryptd
glue_helper fb_sys_fops pcspkr ipmi_msghandler i2c_i801 drm mei_me sg
i2c_smbus lpc_ich mei mfd_core i2c_core ioatdma wmi acpi_power_meter
acpi_pad ip_tables ext4 mbcache jbd2 sd_mod t10_pi crc32c_intel ixgbe
mdio ahci ptp libahci pps_core dca libata
[ 61.431901] CPU: 0 PID: 155 Comm: kworker/0:2 Tainted: G S
5.10.0-rc1+ #19
[ 61.441572] Hardware name: Intel Corporation S2600WTT/S2600WTT, BIOS
SE5C610.86B.01.01.0018.C4.072020161249 07/20/2016
[ 61.454087] Workqueue: events work_for_cpu_fn
[ 61.459521] RIP: 0010:dma_map_page_attrs+0x145/0x1d0
[ 61.465637] Code: 1c 25 28 00 00 00 0f 85 97 00 00 00 48 83 c4 10 5b
5d 41 5c 41 5d c3 4c 89 da eb d7 48 89 f2 48 2b 50 18 48 89 d0 eb 94 0f
0b <0f> 0b 48 c7 c0 ff ff ff ff eb c3 48 89 d9 48 8b 40 40 e8 34 95 af
[ 61.487826] RSP: 0018:ffffc90006a23b70 EFLAGS: 00010246
[ 61.494274] RAX: ffffffff81e25280 RBX: 0000000000000828 RCX:
0000000000000000
[ 61.502874] RDX: 00000000000000b8 RSI: ffffea0004113e00 RDI:
ffff8881152104e8
[ 61.511487] RBP: ffff8881044f8000 R08: 0000000000000002 R09:
0000000000000000
[ 61.520111] R10: 0000000000000002 R11: ffff8881044f8000 R12:
ffff888106847870
[ 61.528743] R13: ffff8881068478a8 R14: ffff8881152104e8 R15:
0000000000000828
[ 61.537387] FS: 0000000000000000(0000) GS:ffff888667a00000(0000)
knlGS:0000000000000000
[ 61.547114] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 61.554230] CR2: 00007f5ed8962248 CR3: 000000000220a004 CR4:
00000000001706f0
[ 61.562913] Call Trace:
[ 61.566393] ib_mad_post_receive_mads+0xd7/0x320 [ib_core]
[ 61.573263] ? _ib_modify_qp+0x310/0x350 [ib_core]
[ 61.579360] ? ib_find_pkey+0x98/0xe0 [ib_core]
[ 61.585174] ib_mad_init_device+0x45f/0x640 [ib_core]
[ 61.591579] add_client_context+0x12b/0x1c0 [ib_core]
[ 61.598020] enable_device_and_get+0xe4/0x1e0 [ib_core]
[ 61.604660] ib_register_device+0x4fb/0x590 [ib_core]
[ 61.611109] ? __vmalloc_node+0x44/0x70
[ 61.616211] rvt_register_device+0x122/0x250 [rdmavt]
[ 61.622737] hfi1_register_ib_device+0x62e/0x6a0 [hfi1]
[ 61.629432] init_one.cold.36+0x284/0x44d [hfi1]
[ 61.635418] local_pci_probe+0x42/0x80
[ 61.640428] work_for_cpu_fn+0x16/0x20
[ 61.645434] process_one_work+0x1aa/0x340
[ 61.650732] ? create_worker+0x1a0/0x1a0
[ 61.655931] worker_thread+0x1cf/0x390
[ 61.660934] ? create_worker+0x1a0/0x1a0
[ 61.666126] kthread+0x116/0x130
[ 61.670531] ? kthread_park+0x80/0x80
[ 61.675420] ret_from_fork+0x22/0x30
[ 61.680202] CPU: 0 PID: 155 Comm: kworker/0:2 Tainted: G S
5.10.0-rc1+ #19
[ 61.690132] Hardware name: Intel Corporation S2600WTT/S2600WTT, BIOS
SE5C610.86B.01.01.0018.C4.072020161249 07/20/2016
[ 61.702892] Workqueue: events work_for_cpu_fn
[ 61.708565] Call Trace:
[ 61.712091] dump_stack+0x6d/0x88
[ 61.716566] __warn.cold.14+0xe/0x3d
[ 61.721309] ? dma_map_page_attrs+0x145/0x1d0
[ 61.726907] report_bug+0xbd/0xf0
[ 61.731319] handle_bug+0x3c/0x90
[ 61.735706] exc_invalid_op+0x13/0x60
[ 61.740460] asm_exc_invalid_op+0x12/0x20
[ 61.745579] RIP: 0010:dma_map_page_attrs+0x145/0x1d0
[ 61.751758] Code: 1c 25 28 00 00 00 0f 85 97 00 00 00 48 83 c4 10 5b
5d 41 5c 41 5d c3 4c 89 da eb d7 48 89 f2 48 2b 50 18 48 89 d0 eb 94 0f
0b <0f> 0b 48 c7 c0 ff ff ff ff eb c3 48 89 d9 48 8b 40 40 e8 34 95 af
[ 61.774015] RSP: 0018:ffffc90006a23b70 EFLAGS: 00010246
[ 61.780496] RAX: ffffffff81e25280 RBX: 0000000000000828 RCX:
0000000000000000
[ 61.789103] RDX: 00000000000000b8 RSI: ffffea0004113e00 RDI:
ffff8881152104e8
[ 61.797689] RBP: ffff8881044f8000 R08: 0000000000000002 R09:
0000000000000000
[ 61.806278] R10: 0000000000000002 R11: ffff8881044f8000 R12:
ffff888106847870
[ 61.814868] R13: ffff8881068478a8 R14: ffff8881152104e8 R15:
0000000000000828
[ 61.823461] ib_mad_post_receive_mads+0xd7/0x320 [ib_core]
[ 61.830214] ? _ib_modify_qp+0x310/0x350 [ib_core]
[ 61.836184] ? ib_find_pkey+0x98/0xe0 [ib_core]
[ 61.841866] ib_mad_init_device+0x45f/0x640 [ib_core]
[ 61.848110] add_client_context+0x12b/0x1c0 [ib_core]
[ 61.854354] enable_device_and_get+0xe4/0x1e0 [ib_core]
[ 61.860789] ib_register_device+0x4fb/0x590 [ib_core]
[ 61.867015] ? __vmalloc_node+0x44/0x70
[ 61.871881] rvt_register_device+0x122/0x250 [rdmavt]
[ 61.878133] hfi1_register_ib_device+0x62e/0x6a0 [hfi1]
[ 61.884585] init_one.cold.36+0x284/0x44d [hfi1]
[ 61.890328] local_pci_probe+0x42/0x80
[ 61.895099] work_for_cpu_fn+0x16/0x20
[ 61.899864] process_one_work+0x1aa/0x340
[ 61.904919] ? create_worker+0x1a0/0x1a0
[ 61.909874] worker_thread+0x1cf/0x390
[ 61.914631] ? create_worker+0x1a0/0x1a0
[ 61.919580] kthread+0x116/0x130
[ 61.923747] ? kthread_park+0x80/0x80
[ 61.928397] ret_from_fork+0x22/0x30
[ 61.932956] ---[ end trace b9195d1a0ae0f872 ]---
-Denny
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: hfi1 broken due to dma_device changes in ib_register
2020-10-29 21:54 hfi1 broken due to dma_device changes in ib_register Dennis Dalessandro
@ 2020-10-29 22:01 ` Jason Gunthorpe
2020-10-29 22:17 ` Bob Pearson
0 siblings, 1 reply; 4+ messages in thread
From: Jason Gunthorpe @ 2020-10-29 22:01 UTC (permalink / raw)
To: Dennis Dalessandro, Bob Pearson; +Cc: linux-rdma
On Thu, Oct 29, 2020 at 05:54:22PM -0400, Dennis Dalessandro wrote:
> Just a heads up, 5.10-rc1 is broken for rdmavt/hfi1 after:
>
> e0477b34d9d11 "(RDMA: Explicitly pass in the dma_device to
> ib_register_device)"
>
> Running with that change causes the call trace below. Reverting the patch
> works around the problem. I haven't yet had a chance to look at what the
> actual cause is, but will and follow up with a proposed patch hopefully
> soon.
Test this:
https://lore.kernel.org/linux-rdma/20201028173108.GA10135@lst.de/T/#mde105a810fb9d2bf734554f3a9875468184dd96c
Jason
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: hfi1 broken due to dma_device changes in ib_register
2020-10-29 22:01 ` Jason Gunthorpe
@ 2020-10-29 22:17 ` Bob Pearson
2020-10-30 7:03 ` Parav Pandit
0 siblings, 1 reply; 4+ messages in thread
From: Bob Pearson @ 2020-10-29 22:17 UTC (permalink / raw)
To: Jason Gunthorpe, Dennis Dalessandro; +Cc: linux-rdma
On 10/29/20 5:01 PM, Jason Gunthorpe wrote:
> On Thu, Oct 29, 2020 at 05:54:22PM -0400, Dennis Dalessandro wrote:
>> Just a heads up, 5.10-rc1 is broken for rdmavt/hfi1 after:
>>
>> e0477b34d9d11 "(RDMA: Explicitly pass in the dma_device to
>> ib_register_device)"
>>
>> Running with that change causes the call trace below. Reverting the patch
>> works around the problem. I haven't yet had a chance to look at what the
>> actual cause is, but will and follow up with a proposed patch hopefully
>> soon.
>
> Test this:
>
> https://lore.kernel.org/linux-rdma/20201028173108.GA10135@lst.de/T/#mde105a810fb9d2bf734554f3a9875468184dd96c
>
> Jason
>
This the same issue I found. I ended up just using DMA_BIT_MASK(64) as was suggested.
^ permalink raw reply [flat|nested] 4+ messages in thread
* RE: hfi1 broken due to dma_device changes in ib_register
2020-10-29 22:17 ` Bob Pearson
@ 2020-10-30 7:03 ` Parav Pandit
0 siblings, 0 replies; 4+ messages in thread
From: Parav Pandit @ 2020-10-30 7:03 UTC (permalink / raw)
To: Bob Pearson, Jason Gunthorpe, Dennis Dalessandro; +Cc: linux-rdma
> From: Bob Pearson <rpearsonhpe@gmail.com>
> Sent: Friday, October 30, 2020 3:48 AM
>
> On 10/29/20 5:01 PM, Jason Gunthorpe wrote:
> > On Thu, Oct 29, 2020 at 05:54:22PM -0400, Dennis Dalessandro wrote:
> >> Just a heads up, 5.10-rc1 is broken for rdmavt/hfi1 after:
> >>
> >> e0477b34d9d11 "(RDMA: Explicitly pass in the dma_device to
> >> ib_register_device)"
> >>
> >> Running with that change causes the call trace below. Reverting the
> >> patch works around the problem. I haven't yet had a chance to look
> >> at what the actual cause is, but will and follow up with a proposed
> >> patch hopefully soon.
> >
> > Test this:
> >
> > https://lore.kernel.org/linux-rdma/20201028173108.GA10135@lst.de/T/#md
> > e105a810fb9d2bf734554f3a9875468184dd96c
> >
> > Jason
> >
> This the same issue I found. I ended up just using DMA_BIT_MASK(64) as was
> suggested.
Ok. sending format fix now. Thanks a lot for the ack and testing.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2020-10-30 7:03 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-10-29 21:54 hfi1 broken due to dma_device changes in ib_register Dennis Dalessandro
2020-10-29 22:01 ` Jason Gunthorpe
2020-10-29 22:17 ` Bob Pearson
2020-10-30 7:03 ` Parav Pandit
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.