From: Guillaume Autran <gautran@mrv•com>
To: Marcelo Tosatti <marcelo.tosatti@cyclades•com>
Cc: "Smith, Craig" <craig.d.smith@siemens•com>,
paulus@samba•org,
linux-ppc-embedded <linuxppc-embedded@ozlabs•org>
Subject: Re: Linux 2.6.x on 8xx status
Date: Wed, 23 Mar 2005 09:06:51 -0500 [thread overview]
Message-ID: <424177FB.7090101@mrv.com> (raw)
In-Reply-To: <20050322175815.GB7846@logos.cnet>
[-- Attachment #1: Type: text/plain, Size: 3299 bytes --]
Marcelo Tosatti wrote:
>On Tue, Mar 22, 2005 at 03:57:08PM -0500, Dan Malek wrote:
>
>
>>On Mar 22, 2005, at 8:04 AM, Marcelo Tosatti wrote:
>>
>>
>>
>>>I'm quite puzzled. Why v2.6 calls the "tlbie" instruction 100-or-so
>>>less times than v2.4 ?
>>>
>>>
>
>That was rather a _factor_ of "100-or-so" less.
>
>
>
>>Oh my ... I'm more worried about the high number of TLB misses
>>in 2.6 compared to 2.4. That's really bad.
>>
>>
>
>Newbie question: What prevents the initial kernel map (tuple of 8Mbyte I/D-TLB entries)
>and the IMMR 8Mbyte D-TLB entry from getting unmapped by translation pressure,
>in case CONFIG_PIN_TLB is disabled ?
>
>
>
>>How did you instrument the tlbie measurement?
>>
>>
>
>By a counter at the end of _tlbie function, similar to other counters which
>you suggested.
>
>
>
>>It could be that 2.4 used lots more 'tlbia' which were replaced by tlbie in 2.6.
>>
>>
>
>Dont think thats the case given that v2.4 calls tlbia through flush_tlb_mm() at exit_mmap()
>only. And at vmalloc_free which shouldnt be called at all.
>
>I just noticed this conditional at switch_mm() (v2.6), which _can_ partly
>explain the reduced tlbie's (its just a guess for now, though):
>
>static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
> struct task_struct *tsk)
>{
>#ifdef CONFIG_ALTIVEC
> asm volatile (
> BEGIN_FTR_SECTION
> "dssall;\n"
>#ifndef CONFIG_POWER4
> "sync;\n" /* G4 needs a sync here, G5 apparently not */
>#endif
> END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
> : : );
>#endif /* CONFIG_ALTIVEC */
>
> tsk->thread.pgdir = next->pgd;
>
> /* No need to flush userspace segments if the mm doesnt change */
> if (prev == next) <--------------
> return; <--------------
>
> /* Setup new userspace context */
> get_mmu_context(next);
> set_context(next->context, next->pgd);
>}
>
>I'm about to disable it and retry.
>
>Spent part of the day reading the MMU section of 860 manual, I think I have kind
>of a clue how things are supposed to work at the lowlevel now.
>
>I'll continue tracking it down - any help is appreciated.
>
>PS: I can't reproduce the invalid TLB crash anymore. i.e. even by removing
>the _tlbie() at update_mmu_cache() everything is working as expected.
>
>How can I reproduce it again? Guillaume, what kernel version are you using?
>
>
>
It is very timing dependant. Running 2.6.11, ldconfig crashes in
__flush_dcache_icache(...) just after boot time (first time called).
Unfortunatly, it only happens every now and then. And of course, never
when my BDI2000 is plugged in. :(
I also noticed that with kernel preemption disable, oops are less
frequent. Probably does not mean anything anyway...
Guillaume.
--
=======================================
Guillaume Autran
Senior Software Engineer
MRV Communications, Inc.
Tel: (978) 952-4932 office
E-mail: gautran@mrv•com
=======================================
[-- Attachment #2: Type: text/html, Size: 4130 bytes --]
next prev parent reply other threads:[~2005-03-23 14:06 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-09-21 11:59 Linux 2.6.x on 8xx status Smith, Craig
2004-09-21 11:53 ` Pantelis Antoniou
2004-09-21 16:54 ` Dan Malek
2004-09-21 17:03 ` Dan Malek
2004-09-22 11:09 ` Sam Song
2004-10-15 15:49 ` Marcelo Tosatti
2004-10-16 2:13 ` Marcelo Tosatti
2004-10-16 9:08 ` Sam Song
2004-10-18 5:59 ` Pantelis Antoniou
2004-10-18 6:22 ` Sam Song
2004-10-18 3:10 ` Dan Malek
2005-02-10 15:04 ` Marcelo Tosatti
2005-02-10 19:26 ` Dan Malek
2005-02-10 17:06 ` Marcelo Tosatti
2005-02-10 17:08 ` Marcelo Tosatti
2005-03-21 21:45 ` Guillaume Autran
2005-03-21 21:53 ` Robert P. J. Day
2005-03-22 13:04 ` Marcelo Tosatti
2005-03-22 20:57 ` Dan Malek
2005-03-22 17:58 ` Marcelo Tosatti
2005-03-22 22:53 ` Dan Malek
2005-03-23 10:25 ` Marcelo Tosatti
2005-03-23 16:12 ` Dan Malek
2005-03-23 16:05 ` Marcelo Tosatti
2005-03-24 14:05 ` Pantelis Antoniou
2005-03-23 14:06 ` Guillaume Autran [this message]
2005-02-11 3:42 ` Dan Malek
2005-02-21 21:12 ` Armin Schindler
2005-02-21 23:45 ` Guillaume Autran
2005-02-15 9:39 ` Reading of RTC robin
-- strict thread matches above, loose matches on Subject: below --
2005-02-22 11:20 Linux 2.6.x on 8xx status Joakim Tjernlund
2004-10-25 15:00 Sam Song
2004-09-17 13:57 Smith, Craig
2004-09-17 13:50 ` Pantelis Antoniou
[not found] <20040916201505.E723B2BDB2@ozlabs.org>
2004-09-17 10:06 ` Song Sam
2004-09-17 9:55 ` Pantelis Antoniou
2004-09-18 20:11 ` Song Sam
2004-09-20 6:02 ` Pantelis Antoniou
2004-09-20 11:47 ` Song Sam
2004-09-20 17:49 ` Tom Rini
2004-09-20 18:02 ` Robert P. J. Day
2004-09-20 18:17 ` Tom Rini
2004-09-21 1:38 ` Song Sam
2004-09-21 6:12 ` Pantelis Antoniou
2004-09-21 10:35 ` Song Sam
2004-09-21 10:41 ` Pantelis Antoniou
2004-10-15 15:47 ` Marcelo Tosatti
2004-09-16 18:56 Smith, Craig
2004-09-16 20:07 ` Dan Malek
2004-09-17 5:43 ` Pantelis Antoniou
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=424177FB.7090101@mrv.com \
--to=gautran@mrv$(echo .)com \
--cc=craig.d.smith@siemens$(echo .)com \
--cc=linuxppc-embedded@ozlabs$(echo .)org \
--cc=marcelo.tosatti@cyclades$(echo .)com \
--cc=paulus@samba$(echo .)org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox