Bug 105968 - ib test on ring 5 succeeded takes too long
Summary: ib test on ring 5 succeeded takes too long
Status: RESOLVED WONTFIX
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Radeon (show other bugs)
Version: DRI git
Hardware: Other All
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-04-10 10:59 UTC by Paul Menzel
Modified: 2018-04-10 11:50 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
Linux 4.16+ messages (112.80 KB, text/plain)
2018-04-10 10:59 UTC, Paul Menzel
no flags Details

Description Paul Menzel 2018-04-10 10:59:35 UTC
Created attachment 138725 [details]
Linux 4.16+ messages

Loading the module *radeon* on the ASRock E350M1, takes 800 ms, which holds up the boot noticeably, and prevents the graphical login to displayed as fast as possible.

I blacklisted the module *radeon*, and loaded it manually with `sudo modprobe radeon`, and it takes 810 ms according to `initcall_debug`.

```
[   97.168677] calling  drm_core_init+0x0/0xc7 [drm] @ 430
[   97.168771] initcall drm_core_init+0x0/0xc7 [drm] returned 0 after 39 usecs
[   97.187351] calling  drm_kms_helper_init+0x0/0x19 [drm_kms_helper] @ 430
[   97.187381] initcall drm_kms_helper_init+0x0/0x19 [drm_kms_helper] returned 0 after 0 usecs
[   97.194638] calling  ttm_init+0x0/0x1000 [ttm] @ 430
[   97.194715] initcall ttm_init+0x0/0x1000 [ttm] returned 0 after 53 usecs
[   97.223316] calling  radeon_init+0x0/0x9c [radeon] @ 430
[   97.223323] [drm] radeon kernel modesetting enabled.
[   97.224106] [drm] initializing kernel modesetting (PALM 0x1002:0x9802 0x1002:0x9802 0x00).
[   97.224203] ATOM BIOS: AMD
[   97.224290] radeon 0000:00:01.0: VRAM: 384M 0x0000000000000000 - 0x0000000017FFFFFF (384M used)
[   97.224295] radeon 0000:00:01.0: GTT: 1024M 0x0000000018000000 - 0x0000000057FFFFFF
[   97.224304] [drm] Detected VRAM RAM=384M, BAR=256M
[   97.224305] [drm] RAM width 32bits DDR
[   97.232611] [TTM] Zone  kernel: Available graphics memory: 412084 kiB
[   97.232617] [TTM] Zone highmem: Available graphics memory: 1596464 kiB
[   97.232618] [TTM] Initializing pool allocator
[   97.232630] [TTM] Initializing DMA pool allocator
[   97.232684] [drm] radeon: 384M of VRAM memory ready
[   97.232687] [drm] radeon: 1024M of GTT memory ready.
[   97.232722] [drm] Loading PALM Microcode
[   97.237441] [drm] Internal thermal controller without fan control
[   97.237694] [drm] Found smc ucode version: 0x00010601
[   97.237802] [drm] radeon: dpm initialized
[   97.239949] [drm] GART: num cpu pages 262144, num gpu pages 262144
[   97.264924] [drm] PCIE GART of 1024M enabled (table at 0x0000000000162000).
[   97.265223] radeon 0000:00:01.0: WB enabled
[   97.265233] radeon 0000:00:01.0: fence driver on ring 0 use gpu addr 0x0000000018000c00 and cpu addr 0x4d2404a7
[   97.265238] radeon 0000:00:01.0: fence driver on ring 3 use gpu addr 0x0000000018000c0c and cpu addr 0xeeb7a49c
[   97.265998] radeon 0000:00:01.0: fence driver on ring 5 use gpu addr 0x0000000000072118 and cpu addr 0x1e8d6f5a
[   97.266002] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
[   97.266004] [drm] Driver supports precise vblank timestamp query.
[   97.266007] radeon 0000:00:01.0: radeon: MSI limited to 32-bit
[   97.266105] radeon 0000:00:01.0: radeon: using MSI.
[   97.266148] [drm] radeon: irq initialized.
[   97.282724] [drm] ring test on 0 succeeded in 1 usecs
[   97.282734] [drm] ring test on 3 succeeded in 3 usecs
[   97.328534] [drm] ring test on 5 succeeded in 1 usecs
[   97.348406] [drm] UVD initialized successfully.
[   97.348698] [drm] ib test on ring 0 succeeded in 0 usecs
[   97.348757] [drm] ib test on ring 3 succeeded in 0 usecs
[   97.888909] [drm] ib test on ring 5 succeeded
[   97.910984] [drm] Radeon Display Connectors
[   97.910993] [drm] Connector 0:
[   97.910995] [drm]   HDMI-A-1
[   97.910998] [drm]   HPD1
[   97.911006] [drm]   DDC: 0x6430 0x6430 0x6434 0x6434 0x6438 0x6438 0x643c 0x643c
[   97.911008] [drm]   Encoders:
[   97.911011] [drm]     DFP1: INTERNAL_UNIPHY
[   97.911014] [drm] Connector 1:
[   97.911016] [drm]   HDMI-A-2
[   97.911018] [drm]   HPD2
[   97.911025] [drm]   DDC: 0x6440 0x6440 0x6444 0x6444 0x6448 0x6448 0x644c 0x644c
[   97.911026] [drm]   Encoders:
[   97.911029] [drm]     DFP2: INTERNAL_UNIPHY
[   97.911032] [drm] Connector 2:
[   97.911034] [drm]   VGA-1
[   97.911040] [drm]   DDC: 0x64d8 0x64d8 0x64dc 0x64dc 0x64e0 0x64e0 0x64e4 0x64e4
[   97.911042] [drm]   Encoders:
[   97.911045] [drm]     CRT1: INTERNAL_KLDSCP_DAC1
[   97.956889] [drm] fb mappable at 0xE0366000
[   97.956895] [drm] vram apper at 0xE0000000
[   97.956896] [drm] size 5242880
[   97.956898] [drm] fb depth is 24
[   97.956899] [drm]    pitch is 5120
[   97.957422] fbcon: radeondrmfb (fb0) is primary device
[   98.015498] Console: switching to colour frame buffer device 160x64
[   98.036405] radeon 0000:00:01.0: fb0: radeondrmfb frame buffer device
[   98.052582] [drm] Initialized radeon 2.50.0 20080528 for 0000:00:01.0 on minor 0
[   98.052896] initcall radeon_init+0x0/0x9c [radeon] returned 0 after 810037 usecs
```
Comment 1 Christian König 2018-04-10 11:25:26 UTC
Yeah, UVD clocking on APUs doesn't work correctly.

We are just sorting that out for the new APUs, but won't have time to fix the old ones as well.
Comment 2 Paul Menzel 2018-04-10 11:50:53 UTC
(In reply to Christian König from comment #1)
> Yeah, UVD clocking on APUs doesn't work correctly.

Interesting. Could you please elaborate? If it does not work correctly anyway, can it be disabled to avoid the boot time penalty?

> We are just sorting that out for the new APUs, but won't have time to fix
> the old ones as well.

Understood.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.