Skip to contentSkip to conversation

New: "sysfs: cannot create duplicate filename /dma/dma0chan0" with 68dbe80f ("crypto: ccp - Release dma channels before dmaengine unrgister")

Bug 216888 Bjorn Helgaas 2 participants 3 messages Jan 5, 2023 – Jan 30, 2023 1 message not in the archive

Subsystems (3)
PCI Crypto API Regressions
lore
Share & tools
lore

Bjorn Helgaas

author
#1 lore raw
In reply to a message not in the archive

Per the report, this is a regression and reverting 68dbe80f5b51 ("crypto: ccp - Release dma channels before dmaengine unrgister"), which appeared in v6.1, avoids the problem.

The bugzilla is assigned to PCI, and I know PCI does have similar sysfs duplicate filename issues, but I don't know whether this instance is related to 68dbe80f5b51 or to the PCI core.

Quoted 101 lines · On Thu, Jan 05, 2023 at 03:12:26PM +0000, [email protected] wrote:

https://bugzilla.kernel.org/show_bug.cgi?id=216888
Summary: "sysfs: cannot create duplicate filename
/dma/dma0chan0" with 68dbe80f ("crypto: ccp - Release dma channels before dmaengine unrgister")

An issue was found in the CCP driver (drivers/crypto/ccp/) that the patch 68dbe80f ("crypto: ccp - Release dma channels before dmaengine unrgister") causes the following errors/warnings with just unloading and loading the CCP driver. I'm not sure if this can be reproduced without the CCP hardware:

# uname -r
6.2.0-0.rc2.18.eln124.x86_64
# lspci | grep 'Encryption controller'
02:00.1 Encryption controller: Advanced Micro Devices, Inc. [AMD]
Zeppelin Cryptographic Coprocessor NTBCCP  ### PCIID: 1022:1468
Subsystem: 1022:1468
03:00.2 Encryption controller: Advanced Micro Devices, Inc. [AMD]
Family 17h (Models 00h-0fh) Platform Security Processor ### PCIID:
1022:1456 Subsystem: 1022:1456
# rmmod kvm_amd ccp ; modprobe ccp
[  140.965403] sysfs: cannot create duplicate filename
'/devices/pci0000:00/0000:00:07.1/0000:03:00.2/dma/dma0chan0'
[  140.975736] CPU: 0 PID: 388 Comm: kworker/0:2 Kdump: loaded Not
tainted 6.2.0-0.rc2.18.eln124.x86_64 #1
[  140.985185] Hardware name: HPE ProLiant DL325 Gen10/ProLiant DL325
Gen10, BIOS A41 07/17/2020
[  140.993761] Workqueue: events work_for_cpu_fn
[  140.998151] Call Trace:
[  141.000613]  <TASK>
[  141.002726]  dump_stack_lvl+0x33/0x46
[  141.006415]  sysfs_warn_dup.cold+0x17/0x23
[  141.010542]  sysfs_create_dir_ns+0xba/0xd0
[  141.014670]  kobject_add_internal+0xba/0x260
[  141.018970]  kobject_add+0x81/0xb0
[  141.022395]  device_add+0xdc/0x7e0
[  141.025822]  ? complete_all+0x20/0x90
[  141.029510]  __dma_async_device_channel_register+0xc9/0x130
[  141.035119]  dma_async_device_register+0x19e/0x3b0
[  141.039943]  ccp_dmaengine_register+0x334/0x3f0 [ccp]
[  141.045042]  ccp5_init+0x662/0x6a0 [ccp]
[  141.049000]  ? devm_kmalloc+0x40/0xd0
[  141.052688]  ccp_dev_init+0xbb/0xf0 [ccp]
[  141.056732]  ? __pci_set_master+0x56/0xd0
[  141.060768]  sp_init+0x70/0x90 [ccp]
[  141.064377]  sp_pci_probe+0x186/0x1b0 [ccp]
[  141.068596]  local_pci_probe+0x41/0x80
[  141.072374]  work_for_cpu_fn+0x16/0x20
[  141.076145]  process_one_work+0x1c8/0x380
[  141.080181]  worker_thread+0x1ab/0x380
[  141.083953]  ? __pfx_worker_thread+0x10/0x10
[  141.088250]  kthread+0xda/0x100
[  141.091413]  ? __pfx_kthread+0x10/0x10
[  141.095185]  ret_from_fork+0x2c/0x50
[  141.098788]  </TASK>
[  141.100996] kobject_add_internal failed for dma0chan0 with -EEXIST,
don't try to register things with the same name in the same directory.
[  141.113703] ccp 0000:03:00.2: ccp initialization failed
[  141.118983] ccp 0000:03:00.2: SEV: memory encryption not enabled by BIOS
[  141.125728] ccp 0000:03:00.2: psp enabled
[  141.130020] ccp 0000:02:00.1: could not enable MSI-X (-22), trying MSI
[  141.136939] sysfs: cannot create duplicate filename '/class/dma/dma0chan0'
[  141.143863] CPU: 0 PID: 388 Comm: kworker/0:2 Kdump: loaded Not
tainted 6.2.0-0.rc2.18.eln124.x86_64 #1
[  141.153313] Hardware name: HPE ProLiant DL325 Gen10/ProLiant DL325
Gen10, BIOS A41 07/17/2020
[  141.161889] Workqueue: events work_for_cpu_fn
[  141.166274] Call Trace:
[  141.168733]  <TASK>
[  141.170845]  dump_stack_lvl+0x33/0x46
[  141.174531]  sysfs_warn_dup.cold+0x17/0x23
[  141.178652]  sysfs_do_create_link_sd+0xcf/0xe0
[  141.183124]  device_add+0x28a/0x7e0
[  141.186634]  ? complete_all+0x20/0x90
[  141.190318]  __dma_async_device_channel_register+0xc9/0x130
[  141.195925]  dma_async_device_register+0x19e/0x3b0
[  141.200745]  ccp_dmaengine_register+0x334/0x3f0 [ccp]
[  141.205838]  ccp5_init+0x662/0x6a0 [ccp]
[  141.209795]  ? devm_kmalloc+0x40/0xd0
[  141.213479]  ccp_dev_init+0xbb/0xf0 [ccp]
[  141.217524]  ? __pci_set_master+0x56/0xd0
[  141.221559]  sp_init+0x70/0x90 [ccp]
[  141.225166]  sp_pci_probe+0x186/0x1b0 [ccp]
[  141.229385]  local_pci_probe+0x41/0x80
[  141.233159]  work_for_cpu_fn+0x16/0x20
[  141.236933]  process_one_work+0x1c8/0x380
[  141.240966]  worker_thread+0x1ab/0x380
[  141.244737]  ? __pfx_worker_thread+0x10/0x10
[  141.249032]  kthread+0xda/0x100
[  141.252192]  ? __pfx_kthread+0x10/0x10
[  141.255964]  ret_from_fork+0x2c/0x50
[  141.259563]  </TASK>
[  141.261902] ccp 0000:02:00.1: ccp initialization failed

This is reproducible with the latest upstream v6.2-rc2 code, "6.2.0-0.rc2.18.eln124.x86_64" is just the v6.2-rc2 code built with Fedora/ELN kernel configs. I'm talking about the commit 68dbe80f because just reverting it makes the above errors/warnings disappear.

#2 lore raw

[TLDR: I'm adding this report to the list of tracked Linux kernel regressions; all text you find below is based on a few templates paragraphs you might have encountered already already in similar form. See link in footer if these mails annoy you.]

Quoted 4 lines · Bjorn Helgaas

Per the report, this is a regression and reverting 68dbe80f5b51 ("crypto: ccp - Release dma channels before dmaengine unrgister"), which appeared in v6.1, avoids the problem.

Quoted 106 lines

The bugzilla is assigned to PCI, and I know PCI does have similar sysfs duplicate filename issues, but I don't know whether this instance is related to 68dbe80f5b51 or to the PCI core.

Quoted 101 lines · On Thu, Jan 05, 2023 at 03:12:26PM +0000, [email protected] wrote:

https://bugzilla.kernel.org/show_bug.cgi?id=216888
Summary: "sysfs: cannot create duplicate filename
/dma/dma0chan0" with 68dbe80f ("crypto: ccp - Release dma channels before dmaengine unrgister")

An issue was found in the CCP driver (drivers/crypto/ccp/) that the patch 68dbe80f ("crypto: ccp - Release dma channels before dmaengine unrgister") causes the following errors/warnings with just unloading and loading the CCP driver. I'm not sure if this can be reproduced without the CCP hardware:

# uname -r
6.2.0-0.rc2.18.eln124.x86_64
# lspci | grep 'Encryption controller'
02:00.1 Encryption controller: Advanced Micro Devices, Inc. [AMD]
Zeppelin Cryptographic Coprocessor NTBCCP  ### PCIID: 1022:1468
Subsystem: 1022:1468
03:00.2 Encryption controller: Advanced Micro Devices, Inc. [AMD]
Family 17h (Models 00h-0fh) Platform Security Processor ### PCIID:
1022:1456 Subsystem: 1022:1456
# rmmod kvm_amd ccp ; modprobe ccp
[  140.965403] sysfs: cannot create duplicate filename
'/devices/pci0000:00/0000:00:07.1/0000:03:00.2/dma/dma0chan0'
[  140.975736] CPU: 0 PID: 388 Comm: kworker/0:2 Kdump: loaded Not
tainted 6.2.0-0.rc2.18.eln124.x86_64 #1
[  140.985185] Hardware name: HPE ProLiant DL325 Gen10/ProLiant DL325
Gen10, BIOS A41 07/17/2020
[  140.993761] Workqueue: events work_for_cpu_fn
[  140.998151] Call Trace:
[  141.000613]  <TASK>
[  141.002726]  dump_stack_lvl+0x33/0x46
[  141.006415]  sysfs_warn_dup.cold+0x17/0x23
[  141.010542]  sysfs_create_dir_ns+0xba/0xd0
[  141.014670]  kobject_add_internal+0xba/0x260
[  141.018970]  kobject_add+0x81/0xb0
[  141.022395]  device_add+0xdc/0x7e0
[  141.025822]  ? complete_all+0x20/0x90
[  141.029510]  __dma_async_device_channel_register+0xc9/0x130
[  141.035119]  dma_async_device_register+0x19e/0x3b0
[  141.039943]  ccp_dmaengine_register+0x334/0x3f0 [ccp]
[  141.045042]  ccp5_init+0x662/0x6a0 [ccp]
[  141.049000]  ? devm_kmalloc+0x40/0xd0
[  141.052688]  ccp_dev_init+0xbb/0xf0 [ccp]
[  141.056732]  ? __pci_set_master+0x56/0xd0
[  141.060768]  sp_init+0x70/0x90 [ccp]
[  141.064377]  sp_pci_probe+0x186/0x1b0 [ccp]
[  141.068596]  local_pci_probe+0x41/0x80
[  141.072374]  work_for_cpu_fn+0x16/0x20
[  141.076145]  process_one_work+0x1c8/0x380
[  141.080181]  worker_thread+0x1ab/0x380
[  141.083953]  ? __pfx_worker_thread+0x10/0x10
[  141.088250]  kthread+0xda/0x100
[  141.091413]  ? __pfx_kthread+0x10/0x10
[  141.095185]  ret_from_fork+0x2c/0x50
[  141.098788]  </TASK>
[  141.100996] kobject_add_internal failed for dma0chan0 with -EEXIST,
don't try to register things with the same name in the same directory.
[  141.113703] ccp 0000:03:00.2: ccp initialization failed
[  141.118983] ccp 0000:03:00.2: SEV: memory encryption not enabled by BIOS
[  141.125728] ccp 0000:03:00.2: psp enabled
[  141.130020] ccp 0000:02:00.1: could not enable MSI-X (-22), trying MSI
[  141.136939] sysfs: cannot create duplicate filename '/class/dma/dma0chan0'
[  141.143863] CPU: 0 PID: 388 Comm: kworker/0:2 Kdump: loaded Not
tainted 6.2.0-0.rc2.18.eln124.x86_64 #1
[  141.153313] Hardware name: HPE ProLiant DL325 Gen10/ProLiant DL325
Gen10, BIOS A41 07/17/2020
[  141.161889] Workqueue: events work_for_cpu_fn
[  141.166274] Call Trace:
[  141.168733]  <TASK>
[  141.170845]  dump_stack_lvl+0x33/0x46
[  141.174531]  sysfs_warn_dup.cold+0x17/0x23
[  141.178652]  sysfs_do_create_link_sd+0xcf/0xe0
[  141.183124]  device_add+0x28a/0x7e0
[  141.186634]  ? complete_all+0x20/0x90
[  141.190318]  __dma_async_device_channel_register+0xc9/0x130
[  141.195925]  dma_async_device_register+0x19e/0x3b0
[  141.200745]  ccp_dmaengine_register+0x334/0x3f0 [ccp]
[  141.205838]  ccp5_init+0x662/0x6a0 [ccp]
[  141.209795]  ? devm_kmalloc+0x40/0xd0
[  141.213479]  ccp_dev_init+0xbb/0xf0 [ccp]
[  141.217524]  ? __pci_set_master+0x56/0xd0
[  141.221559]  sp_init+0x70/0x90 [ccp]
[  141.225166]  sp_pci_probe+0x186/0x1b0 [ccp]
[  141.229385]  local_pci_probe+0x41/0x80
[  141.233159]  work_for_cpu_fn+0x16/0x20
[  141.236933]  process_one_work+0x1c8/0x380
[  141.240966]  worker_thread+0x1ab/0x380
[  141.244737]  ? __pfx_worker_thread+0x10/0x10
[  141.249032]  kthread+0xda/0x100
[  141.252192]  ? __pfx_kthread+0x10/0x10
[  141.255964]  ret_from_fork+0x2c/0x50
[  141.259563]  </TASK>
[  141.261902] ccp 0000:02:00.1: ccp initialization failed

This is reproducible with the latest upstream v6.2-rc2 code, "6.2.0-0.rc2.18.eln124.x86_64" is just the v6.2-rc2 code built with Fedora/ELN kernel configs. I'm talking about the commit 68dbe80f because just reverting it makes the above errors/warnings disappear.

Thanks for the report. To be sure the issue doesn't fall through the cracks unnoticed, I'm adding it to regzbot, the Linux kernel regression tracking bot:

#regzbot introduced 68dbe80f5b51 ^
https://bugzilla.kernel.org/show_bug.cgi?id=216888
#regzbot title cannot create duplicate filename /dma/dma0chan0 #regzbot ignore-activity

This isn't a regression? This issue or a fix for it are already discussed somewhere else? It was fixed already? You want to clarify when the regression started to happen? Or point out I got the title or something else totally wrong? Then just reply and tell me -- ideally while also telling regzbot about it, as explained by the page listed in the footer of this mail.

Reminder for developers: When fixing the issue, add 'Link:' tags pointing to the report (see page linked in footer for details).

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat) --
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
That page also explains what to do if mails like this annoy you.

#3 lore raw

[TLDR: This mail in primarily relevant for Linux kernel regression tracking. See link in footer if these mails annoy you.]

Quoted 11 lines · Linux kernel regression tracking (#adding)
Quoted 10 lines · Bjorn Helgaas

Per the report, this is a regression and reverting 68dbe80f5b51 ("crypto: ccp - Release dma channels before dmaengine unrgister"), which appeared in v6.1, avoids the problem.

The bugzilla is assigned to PCI, and I know PCI does have similar sysfs duplicate filename issues, but I don't know whether this instance is related to 68dbe80f5b51 or to the PCI core.

Quoted 1 line · On Thu, Jan 05, 2023 at 03:12:26PM +0000, [email protected] wrote:

https://bugzilla.kernel.org/show_bug.cgi?id=216888

#regzbot fix: crypto: ccp - Failure on re-initialization due to duplicate sysfs filename

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat) --
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
That page also explains what to do if mails like this annoy you.

#regzbot ignore-activity