Re: [9fans] lock loop messages (anything particular I should do?)

[David Presotto <presotto@closedmind.org>]

[-- Attachment #1: Type: text/plain, Size: 253 bytes --]

Looks to me like lock(&mmukmaplock) should really be an ilock since
mmukmapsync() gets called by trap (i.e. in an interrupt) and by mmukmap()
called in upamalloc() (not necessarily in an interrupt).  I'll change it
and update.  Thanks for the report.

[-- Attachment #2: Type: message/rfc822, Size: 3633 bytes --]

From: Axel Belinfante <Axel.Belinfante@cs.utwente.nl>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] lock loop messages (anything particular I should do?)
Date: Sat, 05 Apr 2003 00:51:36 +0200
Message-ID: <200304042251.h34Mpb502659@zamenhof.cs.utwente.nl>

> > lock 0x8029bd68 loop key 0xdeaddead pc 0x80104e08 held by pc 0x80104e08 pro
> c 486
>
> The numbers are irrelevant to me.  You have to look at them through
> your kernel.

Ok. I was wondering wether it meant something that the same pc number
appeared twice in the lock line. Or is it just one number that gets
printed twice?


Acid tells me:

term% acid 9pcdisk
9pcdisk:386 plan 9 boot image

/sys/lib/acid/port
/sys/lib/acid/386
acid: src(0x80104e08)
/sys/src/9/pc/mmu.c:332
 327
 328		mach0 = MACHP(0);
 329
 330		lock(&mmukmaplock);
 331
>332		if((pte = mmuwalk(mach0->pdb, va, 1, 0)) == nil){
 333			unlock(&mmukmaplock);
 334			return 0;
 335		}
 336		if(!(*pte & PTESIZE) && mmuwalk(mach0->pdb, va, 2, 0) == nil){
 337			unlock(&mmukmaplock);
acid:

My /sys/src/9/pc/mmu.c is of Apr 11 2002.
In case it helpful: I compiled that kernel on Feb 18 2003.


Axel.