Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 05 Oct 2000 19:34:14 -0700 (PDT)
From:      John Baldwin <jhb@FreeBSD.org>
To:        cvs-committers@FreeBSD.org
Cc:        cvs-all@FreeBSD.org
Subject:   RE: cvs commit: src/sys/alpha/alpha trap.c src/sys/alpha/include
Message-ID:  <XFMail.001005193414.jhb@FreeBSD.org>
In-Reply-To: <200010060220.TAA72199@freefall.freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help

On 06-Oct-00 John Baldwin wrote:
> jhb         2000/10/05 19:20:22 PDT
> 
>   Modified files:
>     sys/alpha/alpha      trap.c 
>     sys/alpha/include    cpu.h globals.h 
>     sys/i386/i386        genassym.c globals.s machdep.c 
>                          mp_machdep.c swtch.s trap.c 
>     sys/i386/include     asnames.h cpu.h globaldata.h globals.h 
>                          lock.h mutex.h smptests.h 
>     sys/i386/isa         apic_vector.s clock.c icu_vector.s 
>     sys/kern             kern_clock.c kern_synch.c 
>     sys/sys              proc.h 
>   Log:
>   - Change fast interrupts on x86 to push a full interrupt frame and to
>     return through doreti to handle ast's.  This is necessary for the
>     clock interrupts to work properly.
>   - Change the clock interrupts on the x86 to be fast instead of threaded.
>     This is needed because both hardclock() and statclock() need to run in
>     the context of the current process, not in a separate thread context.

...

This should fix most (if not all) of the micruptime() problems reported after
the SMPng commit.  The P_*PEND trick for delaying signal delivery during
hardclock() was an idea that I gleaned from cp, but don't blame him if I
screwed up the implementation or misunderstood him.  The changes to make
statclock be fast were almost committed with the original SMPng commit instead
of the prevproc hack, albeit in a slightly different form and were submitted
by jake.  Nevertheless, they significantly reduced the work to get statclock
working as most of the code I take straight from jake's patch.

>   Submitted by:       jakeb (making statclock safe in a fast interrupt)
                        ^^^^^

Grr, should be just 'jake'.  IRC nicks != freefall logins. :-P

>   Submitted by:       cp (concept of delaying signals until ast())

-- 

John Baldwin <jhb@FreeBSD.org> -- http://www.FreeBSD.org/~jhb/
PGP Key: http://www.Baldwin.cx/~john/pgpkey.asc
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.001005193414.jhb>