Red Hat Bugzilla – Bug 504950
[965GM] Massive video performance regression with F12 kernel and F11 packages
Last modified: 2013-04-02 00:23:20 EDT
Description of problem:
I was feeling very adventurous so I got the latest 2.6.30 rc from koji and installed it on my brand new f11 setup. I used glxgears as a high level video performance indicator and then use some avi videos and youtube for comparison
This is a thinkpad R61 with intel GM965 and Core 2. running F11 64-bit GNOME.
700 FPS with 2.6.29-4 (the one that came with F11 stock - .167.f11?)
200 FPS with 2.6.30-0.100.rc8.git6.fc12.x86_64!!
[ravis@tenken ~]$ glxgears
1381 frames in 5.0 seconds = 275.892 FPS
1266 frames in 5.0 seconds = 253.128 FPS
FYI - I ran a Debian 5 livecd and it can push 1000+ FPS. but we all know how the intel driver is borked for the time being and will take time to stabilize...here's hoping they fix it soon. Fedora's patched kernels should not add on the already existing poor performance - just my 2c
abysmal video performance. stuttering, stalling and unusable
BTW - should i be even reporting koji build bugs here??
Version-Release number of selected component (if applicable):
Get latest 2.6.30 rc from koji. install and watch videos/youtube/3D apps
Steps to Reproduce:
poor video and 3d performance. I am assuming kernel caused the regression since the intel video driver version is the same as in the old version. all other versions remain the same from F11 stock
improved or equal performance as in F11 stock or older releases
upgraded to kernel-PAE-devel-18.104.22.168-43.fc11.i686.
The video performance is very bad. Made my eyes bleed.
Dell d820 Fedora 11
01:00.0 VGA compatible controller: nVidia Corporation G72M [GeForce Go 7400] (rev a1) (prog-if 00 [VGA controller])
Subsystem: Dell Device 01cc
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
Interrupt: pin A routed to IRQ 16
Region 0: Memory at ed000000 (32-bit, non-prefetchable) [size=16M]
Region 1: Memory at d0000000 (64-bit, prefetchable) [size=256M]
Region 3: Memory at ee000000 (64-bit, non-prefetchable) [size=16M]
[virtual] Expansion ROM at ef000000 [disabled] [size=128K]
Capabilities:  Power Management version 2
Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
Capabilities:  MSI: Enable- Count=1/1 Maskable- 64bit+
Address: 0000000000000000 Data: 0000
Capabilities:  Express (v1) Endpoint, MSI 00
DevCap: MaxPayload 128 bytes, PhantFunc 0, Latency L0s <256ns, L1 <4us
ExtTag- AttnBtn- AttnInd- PwrInd- RBE- FLReset-
DevCtl: Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
RlxdOrd+ ExtTag- PhantFunc- AuxPwr- NoSnoop+
MaxPayload 128 bytes, MaxReadReq 512 bytes
DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
LnkCap: Port #0, Speed 2.5GT/s, Width x16, ASPM L0s L1, Latency L0 <256ns, L1 <4us
ClockPM- Surprise- LLActRep- BwNot-
LnkCtl: ASPM Disabled; RCB 128 bytes Disabled- Retrain- CommClk+
ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
LnkSta: Speed 2.5GT/s, Width x16, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
Capabilities:  Virtual Channel <?>
Capabilities:  Power Budgeting <?>
Kernel driver in use: nvidia
Kernel modules: nvidia, nouveau
Brian, this is not the same issue as original report. Please complain to NVIDIA about issues with their proprietary drivers since Fedora can't address them. If you used nouveau, that would be another situation.
Ravi, how about _complete_ Fedora 12 Beta (you can try with LiveCD)?
Ok, thought as much.
They fixed it in the 190.42 release. I just thought that since the nouveau & nvidia drivers showed the same problem at the same time that there might be a related issue. I do not believe in coincidences.
This has been fixed in current post-Fedora 12 Beta rawhide packages.
Intel driver is actually working quite nice now.