Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-05-03 Thread Henri Verbeet
On 3 May 2012 07:17, Alexey Loukianov mooro...@mail.ru wrote: Trying to pinpoint the cause using oprofile produced no valuable results: it either me not able to use this wonderful profiler correctly or the issue is of such kind that isn't easily tracked by oprofile. Personally I think perf is

Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-05-02 Thread Alexey Loukianov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 15.04.2012 07:22, Alexey Loukianov wrote: Good day to all. Henri, Stefan, I address this message to you at the first place as to a main developers of wined3d/opengl stuff... ---8--- strip ---8--- ... an app (localized version of Perfect World

Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-04-16 Thread Roderick Colenbrander
On 4/16/12, Alexey Loukianov mooro...@mail.ru wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 16.04.2012 04:28, Vitaliy Margolen wrote: Of course it won't - they are binary blobs from Nvidia. Not much to see there. All you really looking for are time spent in that library. Vitaliy, I

Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-04-15 Thread Stefan Dösinger
Am Sonntag, 15. April 2012, 07:22:34 schrieb Alexey Loukianov: When I configure an app to run in a windowed mode I've got around 40 FPS on game login screen with nVIDIA drivers 275.09.07, but switching into using more recent versions causes FPS to drop to around ~10. It could be a driver bug as

Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-04-15 Thread Alexey Loukianov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 15.04.2012 21:50, Stefan Dösinger wrote: It could also be because of some additional features added in newer drivers. 16 byte alignment for vertex buffers is a possibility, I believe it was added in the 280 drivers. You can check this by disabling

Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-04-15 Thread Alexey Loukianov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 15.04.2012 21:50, Stefan Dösinger wrote: your best bet would be using something like oprofile to find out which GL calls show performance changes. Well, I had compiled/installed APITrace 3.0 and oprofile 0.9.7 on my system, but it seems that it'd

Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-04-15 Thread Vitaliy Margolen
On 04/15/2012 04:44 PM, Alexey Loukianov wrote: With oprofile I hit another trouble - it seems that this tool is unable to fetch symbols from libGL Of course it won't - they are binary blobs from Nvidia. Not much to see there. All you really looking for are time spent in that library.

Re: Request for help/advice in investigation of one interesting huge FPS regression bug

2012-04-15 Thread Alexey Loukianov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 16.04.2012 04:28, Vitaliy Margolen wrote: Of course it won't - they are binary blobs from Nvidia. Not much to see there. All you really looking for are time spent in that library. Vitaliy, I don't expect oprofile to find hidden COFF or DWARF 2

Request for help/advice in investigation of one interesting huge FPS regression bug

2012-04-14 Thread Alexey Loukianov
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Good day to all. Henri, Stefan, I address this message to you at the first place as to a main developers of wined3d/opengl stuff. Nevertheless, hints and help are welcome from anyone, cause ATM I'm totally confused and don't know what else to try to