Ok, thanks.
On Thu, May 22, 2008 at 8:18 PM, Corey J Ashford <[EMAIL PROTECTED]> wrote: > To be clear, I meant I had held off "creating and posting" a patch for this > change. I will create and post a patch by the end of the day. > > - Corey > > "stephane eranian" <[EMAIL PROTECTED]> wrote on 05/22/2008 10:58:43 AM: >> Corey, >> >> Please post the patch. >> >> On Thu, May 22, 2008 at 7:30 PM, Corey J Ashford <[EMAIL PROTECTED]> >> wrote: >> > Thanks Stephane. I have no objection to removing those macros, but held >> > off >> > posting a patch for that because I remembered that Phil Mucci (as >> > discussed) >> > wanted to keep them. >> > >> > Regards, >> > >> > - Corey >> > >> > "stephane eranian" <[EMAIL PROTECTED]> wrote on 05/22/2008 02:05:59 >> > AM: >> > >> >> Corey, >> >> >> >> Patch applied. >> >> >> >> I would like to remove the pfm_spin_*lock() macros. They don't have >> >> any current >> >> use and that would simplify the code and make it more readable. >> >> Moreover, it won't >> >> give reviewers another argument for saying perfmon is bloated. >> >> >> >> thanks. >> >> >> >> >> >> On Wed, May 21, 2008 at 2:40 AM, Corey Ashford <[EMAIL PROTECTED]> >> >> wrote: >> >> > Hello, >> >> > >> >> > Attached is a patch fixes a problem with the "lazy [aka soft] >> >> > interrupt >> >> > disabling" mechanism in the perfmon2 port to the Power architecture. >> >> > My >> >> > original solution of using wrapper macros for the spin_*lock* calls >> >> > was >> >> > insufficient because interrupts needed to be locked out in more >> >> > places >> >> > than >> >> > perfmon2 itself (e.g. sched()). The solution in this patch is to >> >> > change >> >> > the >> >> > behavior of the interrupt handler so that it sets a cpu-specific >> >> > flag, >> >> > clears the PMU interrupt, disables hardware interrupts and returns. >> >> > When >> >> > interrupts are reenabled again, the flag is checked and the PMU >> >> > interrupt is >> >> > set again. The special wrapper macros are no longer used. >> >> > >> >> > The upside of this change is that it doesn't require any changes to >> >> > the >> >> > head_64.S file to change the wrapper used for the PMU exception, and >> >> > should >> >> > also fix a potential problem with PMU interrupts being lost (I might >> >> > have >> >> > seen this problem before, but I'm not positive). >> >> > >> >> > With this patch, there is still a problem with a kernel hang in a >> >> > test >> >> > case >> >> > I have that generates about 2000 interrupts per second, which then >> >> > signal a >> >> > user-space thread. However, the behavior with this patch is improved >> >> > (doesn't hang as often) and I'm no longer seeing any problems with >> >> > spin >> >> > locks. >> >> > >> >> > Please review it and let me know if you see any problems. >> >> > >> >> > Thanks for your consideration, >> >> > >> >> > - Corey >> >> > >> >> > -- >> >> > Corey Ashford >> >> > Software Engineer >> >> > IBM Linux Technology Center, Linux Toolchain >> >> > Beaverton, OR >> >> > 503-578-3507 >> >> > [EMAIL PROTECTED] >> >> > >> >> > >> >> > >> >> > ------------------------------------------------------------------------- >> >> > This SF.net email is sponsored by: Microsoft >> >> > Defy all challenges. Microsoft(R) Visual Studio 2008. >> >> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >> >> > _______________________________________________ >> >> > perfmon2-devel mailing list >> >> > perfmon2-devel@lists.sourceforge.net >> >> > https://lists.sourceforge.net/lists/listinfo/perfmon2-devel >> >> > >> >> > >> > > ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ perfmon2-devel mailing list perfmon2-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
