First "glitch1" results, 2.6.21-rc7-git6-CFSv5
I am not sure a binary attachment will go thru, I will move to the web ste if not.
Re: [REPORT] First "glitch1" results, 2.6.21-rc7-git6-CFSv5 + SD 0.46
Quoted 2 lines · Bill Davidsen
I am not sure a binary attachment will go thru, I will move to the web ste if not.
I did a quick try of this script here.
With SD 0.46 with X at nice 0 I was getting 1-2 frames per second. I decided to try cfs v5. The option disable auto renicing did not work so many threads other than X are now at -19...
SD 0.46 1-2 FPS
cfs v5 nice -19 219-233 FPS
cfs v5 nice 0 1000-1996
Looks like, in this case, nice -19 for X is NOT a good idea.
Kernel is 2.6.20.7 (gentoo) UP amd64 with HZ 300 voluntary prempt (a fully premptable kernel eventually locks up switching between 32 and 64 apps)
Thanks,
Ed Tomlinson
Re: [REPORT] First "glitch1" results, 2.6.21-rc7-git6-CFSv5 + SD 0.46
Quoted 4 lines · Ed Tomlinson
Quoted 2 lines · Bill Davidsen
I am not sure a binary attachment will go thru, I will move to the web ste if not.
Quoted 8 lines
I did a quick try of this script here.
With SD 0.46 with X at nice 0 I was getting 1-2 frames per second. I decided to try cfs v5. The option disable auto renicing did not work so many threads other than X are now at -19...
SD 0.46 1-2 FPS
cfs v5 nice -19 219-233 FPS
cfs v5 nice 0 1000-1996
cfs v5 nice -10 60-65 FPS
Quoted 5 lines
Looks like, in this case, nice -19 for X is NOT a good idea.
Kernel is 2.6.20.7 (gentoo) UP amd64 with HZ 300 voluntary prempt (a fully premptable kernel eventually locks up switching between 32 and 64 apps)
Thanks
Ed
Re: [REPORT] First "glitch1" results, 2.6.21-rc7-git6-CFSv5 + SD 0.46
Quoted 4 lines · Ed Tomlinson
Quoted 3 lines
SD 0.46 1-2 FPS
cfs v5 nice -19 219-233 FPS
cfs v5 nice 0 1000-1996cfs v5 nice -10 60-65 FPS
the problem is, the glxgears portion of this test is an _inverse_ testcase.
The reason? glxgears on true 3D hardware will _not_ use X, it will directly use the 3D driver of the kernel. So by renicing X to -19 you give the xterms more chance to show stuff - the performance of the glxgears will 'degrade' - but that is what you asked for: glxgears is 'just another CPU hog' that competes with X, it's not a "true" X client.
if you are after glxgears performance in this test then you'll get the best performance out of this by renicing X to +19 or even SCHED_BATCH.
Ingo
Re: [REPORT] First "glitch1" results, 2.6.21-rc7-git6-CFSv5 + SD 0.46
Quoted 5 lines · Ingo Molnar
Quoted 3 lines · Ed Tomlinson
Quoted 3 lines
SD 0.46 1-2 FPS
cfs v5 nice -19 219-233 FPS
cfs v5 nice 0 1000-1996
Quoted 8 lines
Quoted 1 line
cfs v5 nice -10 60-65 FPSthe problem is, the glxgears portion of this test is an _inverse_ testcase.
The reason? glxgears on true 3D hardware will _not_ use X, it will directly use the 3D driver of the kernel. So by renicing X to -19 you give the xterms more chance to show stuff - the performance of the glxgears will 'degrade' - but that is what you asked for: glxgears is 'just another CPU hog' that competes with X, it's not a "true" X client.
if you are after glxgears performance in this test then you'll get the best performance out of this by renicing X to +19 or even
SCHED_BATCH.
Several points on this...
First, I don't think this is accelerated in the way you mean, the machine is a test server, with motherboard video using the 945G video driver. Given the limitations of the support in that setup, I don't think it qualified as "true 3D hardware," although I guess I could try using the vesafb version as a test.
The 2nd thing I note is that on FC6 this scheduler seems to confuse 'top' to some degree, since the glxgears is shown as taking 51% of the CPU (one core), while the state breakdown shows about 73% in idle, waitio, and int. image attached.
After I upgrade the kernel and cfs to the absolute latest I'll repeat this, as well as test with vesafb, and my planned run under heavy load.
Signature · 3 lines
Bill Davidsen <[email protected]> "We have more to fear from the bungling of the incompetent than from the machinations of the wicked." - from Slashdot
Re: [REPORT] First "glitch1" results, 2.6.21-rc7-git6-CFSv5 + SD 0.46
Quoted 28 lines · Bill Davidsen
Quoted 10 lines · Ingo Molnar
Quoted 5 lines · Ed Tomlinson
Quoted 3 lines
SD 0.46 1-2 FPS
cfs v5 nice -19 219-233 FPS
cfs v5 nice 0 1000-1996cfs v5 nice -10 60-65 FPSthe problem is, the glxgears portion of this test is an _inverse_ testcase.
Quoted 8 lines
The reason? glxgears on true 3D hardware will _not_ use X, it will directly use the 3D driver of the kernel. So by renicing X to -19 you give the xterms more chance to show stuff - the performance of the glxgears will 'degrade' - but that is what you asked for: glxgears is 'just another CPU hog' that competes with X, it's not a "true" X client.
if you are after glxgears performance in this test then you'll get the best performance out of this by renicing X to +19 or even
SCHED_BATCH.Several points on this...
First, I don't think this is accelerated in the way you mean, the machine is a test server, with motherboard video using the 945G video driver. Given the limitations of the support in that setup, I don't think it qualified as "true 3D hardware," although I guess I could try using the vesafb version as a test.
Quoted 4 lines
The 2nd thing I note is that on FC6 this scheduler seems to confuse 'top' to some degree, since the glxgears is shown as taking 51% of the CPU (one core), while the state breakdown shows about 73% in idle, waitio, and int. image attached.
top by itself certainly cannot be trusted to give true representation of the cpu usage I'm afraid. It's not as convoluted as, say, trying to track memory usage of an application, but top's resolution being tied to HZ accounting makes it not reliable in that regard.
Quoted 3 lines
After I upgrade the kernel and cfs to the absolute latest I'll repeat this, as well as test with vesafb, and my planned run under heavy load.
I have a problem with your test case Bill. Its behaviour would depend on how gpu bound vs cpu bound vs accelerated vs non-accelerated your graphics card is. I get completely different results to those of the other testers given the different hardware configuration and I don't think my results are valuable. My problem with this testcase is - What would you define as "perfect" behaviour for your test case? It seems far too arbitrary.
Signature · 1 line
-ck
Re: [REPORT] First "glitch1" results, 2.6.21-rc7-git6-CFSv5 + SD 0.46
Quoted 42 lines · Con Kolivas
Quoted 28 lines · Bill Davidsen
Quoted 10 lines · Ingo Molnar
Quoted 5 lines · Ed Tomlinson
Quoted 3 lines
SD 0.46 1-2 FPS
cfs v5 nice -19 219-233 FPS
cfs v5 nice 0 1000-1996cfs v5 nice -10 60-65 FPSthe problem is, the glxgears portion of this test is an _inverse_ testcase.
Quoted 8 lines
The reason? glxgears on true 3D hardware will _not_ use X, it will directly use the 3D driver of the kernel. So by renicing X to -19 you give the xterms more chance to show stuff - the performance of the glxgears will 'degrade' - but that is what you asked for: glxgears is 'just another CPU hog' that competes with X, it's not a "true" X client.
if you are after glxgears performance in this test then you'll get the best performance out of this by renicing X to +19 or even
SCHED_BATCH.Several points on this...
First, I don't think this is accelerated in the way you mean, the machine is a test server, with motherboard video using the 945G video driver. Given the limitations of the support in that setup, I don't think it qualified as "true 3D hardware," although I guess I could try using the vesafb version as a test.
Quoted 4 lines
The 2nd thing I note is that on FC6 this scheduler seems to confuse 'top' to some degree, since the glxgears is shown as taking 51% of the CPU (one core), while the state breakdown shows about 73% in idle, waitio, and int. image attached.
top by itself certainly cannot be trusted to give true representation of the cpu usage I'm afraid. It's not as convoluted as, say, trying to track memory usage of an application, but top's resolution being tied to HZ accounting makes it not reliable in that regard.
Quoted 3 lines
After I upgrade the kernel and cfs to the absolute latest I'll repeat this, as well as test with vesafb, and my planned run under heavy load.
Quoted 6 lines
I have a problem with your test case Bill. Its behaviour would depend on how gpu bound vs cpu bound vs accelerated vs non-accelerated your graphics card is. I get completely different results to those of the other testers given the different hardware configuration and I don't think my results are valuable. My problem with this testcase is - What would you define as "perfect" behaviour for your test case? It seems far too arbitrary.
Con,
One thing I did not mention in all this is that renicing the glxgears process to -10 gets SD to give about 1000FPS, indeed you get most of this performance at -5 too. All in all SD does a very good job here.
Get well soon!
Ed
Re: [REPORT] First "glitch1" results, 2.6.21-rc7-git6-CFSv5 + SD 0.46
In reply to Con KolivasQuoted 38 lines · Con Kolivas
Quoted 30 lines · Bill Davidsen
Quoted 11 lines · Ingo Molnar
Quoted 6 lines · Ed Tomlinson
Quoted 4 lines
SD 0.46 1-2 FPS
cfs v5 nice -19 219-233 FPS
cfs v5 nice 0 1000-1996cfs v5 nice -10 60-65 FPS
the problem is, the glxgears portion of this test is an _inverse_ testcase.
Quoted 9 lines
The reason? glxgears on true 3D hardware will _not_ use X, it will directly use the 3D driver of the kernel. So by renicing X to -19 you give the xterms more chance to show stuff - the performance of the glxgears will 'degrade' - but that is what you asked for: glxgears is 'just another CPU hog' that competes with X, it's not a "true" X client.
if you are after glxgears performance in this test then you'll get the best performance out of this by renicing X to +19 or even
SCHED_BATCH.Several points on this...
First, I don't think this is accelerated in the way you mean, the machine is a test server, with motherboard video using the 945G video driver. Given the limitations of the support in that setup, I don't think it qualified as "true 3D hardware," although I guess I could try using the vesafb version as a test.
Quoted 5 lines
The 2nd thing I note is that on FC6 this scheduler seems to confuse 'top' to some degree, since the glxgears is shown as taking 51% of the CPU (one core), while the state breakdown shows about 73% in idle, waitio, and int. image attached.
Quoted 9 lines
top by itself certainly cannot be trusted to give true representation of the cpu usage I'm afraid. It's not as convoluted as, say, trying to track memory usage of an application, but top's resolution being tied to HZ accounting makes it not reliable in that regard.
Quoted 3 lines
After I upgrade the kernel and cfs to the absolute latest I'll repeat this, as well as test with vesafb, and my planned run under heavy load.
I have a problem with your test case Bill. Its behaviour would depend on how gpu bound vs cpu bound vs accelerated vs non-accelerated your graphics card is. I get completely different results to those of the other testers given the different hardware configuration and I don't think my results are valuable. My problem with this testcase is - What would you define as "perfect" behaviour for your test case? It seems far too arbitrary.
It was more intended to give an immediate feedback on gross behavior. On some old schedulers (2.4.x) it visibly ran one xterm after the other, while on 2.6.2[01] that behavior is gone and all schedulers give equal time as seen by the eye. Looking at the behavior with line and jump scroll, under load or not, X nice or nasty, allows a quick check on where the bad cases are if any exist.
I intended it as a quick way to determine really, visibly, bad scheduling, not a a test for quantifying performance. The fact that fps varies by almost an order of magnitude with some earlier versions of the schedulers is certainly a red flag to me that there's a corner case, and something I care more about than glxgears will be inconsistent as well.
Hopefully in that context, as a relatively quick way to try nice and load values, it's a useful tool.
Signature · 3 lines
bill davidsen <[email protected]> CTO TMR Associates, Inc Doing interesting things with small computers since 1979