<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"> <head> <meta http-equiv="Content-Type" content="text/html; charset=us-ascii"> <meta name="Generator" content="Microsoft Word 15 (filtered medium)"> <style></style> </head> <body lang="EN-US" link="#0563C1" vlink="#954F72" style="word-wrap:break-word"> <div class="WordSection1"> Hello Vinschen,<o:p></o:p> <o:p> </o:p> Hope you are doing well. I am Chaitanya from the linux graphics team in Intel.<o:p></o:p> <o:p> </o:p> This mail is regarding a regression we are seeing in our CI runs[1] on drm-tip[2] repository.<o:p></o:p> <o:p> </o:p> Since the version CI_DRM_14817[3], we are seeing the following regression<o:p></o:p> <o:p> </o:p> `````````````````````````````````````````````````````````````````````````````````<o:p></o:p> 4>[ 10.361988] Hardware name: Intel Corporation Arrow Lake Client Platform/MTL-S UDIMM 2DPC EVCRB, BIOS MTLSFWI1.R00.3473.D80.2311222130 11/22/2023<o:p></o:p> <4>[ 10.362613] Call Trace:<o:p></o:p> <4>[ 10.363227] <TASK><o:p></o:p> <4>[ 10.363835] dump_stack_lvl+0x82/0xd0<o:p></o:p> <4>[ 10.364446] register_lock_class+0x795/0x7e0<o:p></o:p> <4>[ 10.365054] ? __free_object+0xa1/0x340<o:p></o:p> <4>[ 10.365659] ? lockdep_hardirqs_on+0xc1/0x140<o:p></o:p> <4>[ 10.366261] __lock_acquire+0x73/0x2230<o:p></o:p> <4>[ 10.366857] ? _raw_spin_unlock_irqrestore+0x58/0x70<o:p></o:p> <4>[ 10.367451] lock_acquire+0xd9/0x2f0<o:p></o:p> <4>[ 10.368038] ? igc_ptp_clear_tx_tstamp+0x28/0x60 [igc]<o:p></o:p> <4>[ 10.368638] _raw_spin_lock_irqsave+0x3d/0x60<o:p></o:p> <4>[ 10.369212] ? igc_ptp_clear_tx_tstamp+0x28/0x60 [igc]<o:p></o:p> <4>[ 10.369794] igc_ptp_clear_tx_tstamp+0x28/0x60 [igc]<o:p></o:p> <4>[ 10.370369] igc_ptp_set_timestamp_mode.isra.0+0x20b/0x230 [igc]<o:p></o:p> <4>[ 10.370944] igc_ptp_reset+0x31/0x180 [igc]<o:p></o:p> <4>[ 10.371513] igc_reset+0xb4/0x100 [igc]<o:p></o:p> <4>[ 10.372083] igc_probe+0x797/0x8e0 [igc]<o:p></o:p> <4>[ 10.372647] pci_device_probe+0x95/0x120<o:p></o:p> <4>[ 10.373199] really_probe+0xd9/0x370<o:p></o:p> <4>[ 10.373748] ? __pfx___driver_attach+0x10/0x10<o:p></o:p> <4>[ 10.374291] __driver_probe_device+0x73/0x150<o:p></o:p> <4>[ 10.374835] driver_probe_device+0x19/0xa0<o:p></o:p> <4>[ 10.375375] __driver_attach+0xb6/0x180<o:p></o:p> <4>[ 10.375908] ? __pfx___driver_attach+0x10/0x10<o:p></o:p> <4>[ 10.376438] bus_for_each_dev+0x77/0xd0<o:p></o:p> <4>[ 10.376968] bus_add_driver+0x110/0x240<o:p></o:p> <4>[ 10.377492] driver_register+0x5b/0x110<o:p></o:p> <4>[ 10.378015] ? __pfx_igc_init_module+0x10/0x10 [igc]<o:p></o:p> <4>[ 10.378551] do_one_initcall+0x5c/0x2b0<o:p></o:p> <4>[ 10.379079] ? kmalloc_trace_noprof+0x22f/0x290<o:p></o:p> <4>[ 10.379602] ? do_init_module+0x1e/0x210<o:p></o:p> <4>[ 10.380131] do_init_module+0x5f/0x210<o:p></o:p> <4>[ 10.380651] load_module+0x1d44/0x1fc0<o:p></o:p> <4>[ 10.381173] ? init_module_from_file+0x86/0xd0<o:p></o:p> <4>[ 10.381691] init_module_from_file+0x86/0xd0<o:p></o:p> <4>[ 10.382206] idempotent_init_module+0x17c/0x230<o:p></o:p> <4>[ 10.382723] __x64_sys_finit_module+0x56/0xb0<o:p></o:p> <4>[ 10.383237] do_syscall_64+0x66/0x140<o:p></o:p> <4>[ 10.383751] entry_SYSCALL_64_after_hwframe+0x76/0x7e<o:p></o:p> <4>[ 10.384264] RIP: 0033:0x7f668aeff95d<o:p></o:p> `````````````````````````````````````````````````````````````````````````````````<o:p></o:p> Details log can be found in [4].<o:p></o:p> <o:p> </o:p> We have identified the following patch [5] to be the first "bad" commit<o:p></o:p> <o:p> </o:p> `````````````````````````````````````````````````````````````````````````````````````````````````````````<o:p></o:p> commit 86167183a17e03ec77198897975e9fdfbd53cb0b<o:p></o:p> Author: Corinna Vinschen <a href="mailto:vinschen@redhat.com"> vinschen@redhat.com</a><o:p></o:p> Date: Tue Apr 23 12:24:54 2024 +0200<o:p></o:p> <o:p> </o:p> igc: fix a log entry using uninitialized netdev<o:p></o:p> <o:p> </o:p> During successful probe, igc logs this:<o:p></o:p> <o:p> </o:p> [ 5.133667] igc 0000:01:00.0 (unnamed net_device) (uninitialized): PHC added<o:p></o:p> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^<o:p></o:p> The reason is that igc_ptp_init() is called very early, even before<o:p></o:p> register_netdev() has been called. So the netdev_info() call works<o:p></o:p> on a partially uninitialized netdev.<o:p></o:p> <o:p> </o:p> Fix this by calling igc_ptp_init() after register_netdev(), right<o:p></o:p> after the media autosense check, just as in igb. Add a comment,<o:p></o:p> just as in igb.<o:p></o:p> <o:p> </o:p> `````````````````````````````````````````````````````````````````````````````````````````````````````````<o:p></o:p> <o:p> </o:p> Could you please check why the patch causes this regression and provide a fix if necessary?<o:p></o:p> <o:p> </o:p> Thank you.<o:p></o:p> <o:p> </o:p> Regards<o:p></o:p> <o:p> </o:p> Chaitanya<o:p></o:p> <o:p> </o:p> [1] <a href="https://intel-gfx-ci.01.org/tree/drm-tip/index.html"> https://intel-gfx-ci.01.org/tree/drm-tip/index.html</a>? [2] <a href="https://cgit.freedesktop.org/drm-tip/">https://cgit.freedesktop.org/drm-tip/</a> [3] <a href="https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14817/git-log-oneline.txt"> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14817/git-log-oneline.txt</a><o:p></o:p> [4] <a href="https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14817/bat-arls-1/boot0.txt"> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14817/bat-arls-1/boot0.txt</a><o:p></o:p> [5] <a href="https://cgit.freedesktop.org/drm-tip/commit/?id=86167183a17e03ec77198897975e9fdfbd53cb0b">https://cgit.freedesktop.org/drm-tip/commit/?id=86167183a17e03ec77198897975e9fdfbd53cb0b</a><o:p></o:p> </div> </body> </html>