I took some latency traces on our sa1100 (see attachment) running 2.6.15-rc7-rt1 with full preemption. They look very bad at the first glance, but I cannot interpret them in detail. Trace 3, 4 and 5 seem to have obvious reasons: FPU emulation, Framebuffer console updates, and compression/decompression of flash data for jffs2. Moreover, if I read these traces correctly, they just disable preemption, but still allow interrupts. Is that correct? Can anything be done against these latencies, i.e. do they really need to disable preemption for such a long time? However, traces 1, 2, 6 and 7 are completely mysterious to me. Interrupts seem to be blocked for milliseconds, while nothing is going on on the system? Moreover, there are console-related function names in traces 6 and 7, although I've unconfigured the framebuffer console for these runs! Many thanks in advance for any help! Klaus Kusche Entwicklung Software - Steuerung Software Development - Control KEBA AG A-4041 Linz Gewerbepark Urfahr Tel +43 / 732 / 7090-3120 Fax +43 / 732 / 7090-6301 E-Mail: [email protected] www.keba.com
Attachment:
traces.tar.bz2
Description: traces.tar.bz2
- Follow-Ups:
- Re: Latency traces I cannot interpret (sa1100, 2.6.15-rc7-rt1)
- From: Lee Revell <[email protected]>
- Re: Latency traces I cannot interpret (sa1100, 2.6.15-rc7-rt1)
- From: Daniel Walker <[email protected]>
- Re: Latency traces I cannot interpret (sa1100, 2.6.15-rc7-rt1)
- Prev by Date: Re: [patch 00/2] improve .text size on gcc 4.0 and newer compilers
- Next by Date: Re: PCI DMA burst delay
- Previous by thread: page refcounts?
- Next by thread: Re: Latency traces I cannot interpret (sa1100, 2.6.15-rc7-rt1)
- Index(es):
 |