On Sat, 1 Jul 2000, Marcelo Tosatti wrote:
>On Sat, 1 Jul 2000, Andrea Arcangeli wrote:
>
>> On Sat, 1 Jul 2000, Andrea Arcangeli wrote:
>>
>> >On Sat, 1 Jul 2000, Marcelo Tosatti wrote:
>> >
>> >>Waiting on dirty buffers to sync.
>> >
>> >Could you elaborate? I never added any additional wait_on_buffer()
>> >anywhere in 2.2.x. I see potential problem with ll_rw_block(WRITE)
>> >(backported from 2.4.x) and I provided a patch to back it out a few days
>> >ago but that wasn't waiting dirty buffers to sync either (it was only
>> >waiting to get a I/O request slot if they was all full).
>Yes. But my patch (backported from 2.4.x too) which is included into
>2.2.17pre9 waits on dirty buffers.
See below:
On Sat, 1 Jul 2000, Andrea Arcangeli wrote:
>
>Date: Sat, 1 Jul 2000 16:01:52 +0200 (CEST)
>From: Andrea Arcangeli <andrea@suse.de>
>To: Rik van Riel <riel@conectiva.com.br>
>Cc: Marcelo Tosatti <marcelo@conectiva.com.br>, Alan Cox <alan@redhat.com>,
> Linux Kernel <linux-kernel@vger.rutgers.edu>
>Subject: Re: [PATCH] 2.2.17pre7 VM enhancement Re: I/O performance on 2.4.0-test2
>
>On Sat, 1 Jul 2000, Rik van Riel wrote:
>
>>On Wed, 28 Jun 2000, Andrea Arcangeli wrote:
>>
>>> >Testing dbench with 48 clients, I got 17MB/s with 2.2.17pre9, and 21MB/sec
>>> >with 2.2.17pre9 + this patch.
>>> ^^^^^^^^^^^^
>>> I spend an relevat effort to take all the patches separated and ordered,
>>> so I'd appreciate if you could avoid to mix them.
>>
>>Andrea, I'd appreciate it if you could quit
>>the obvious flamebaits and start looking at
>>2.2 VM in a serious way.
>
>???
>
>>Also, 2.2 is supposed to be a stable kernel.
>>We cannot make any tradeoff where we lose some
>>stability for a performance gain.
>
>Which patch are you referring about when you say "lose some stability"?
>
>Andrea
>
>
Rik as far as I understand english was complaining something I did
claiming I caused loss of stability (in the contect of the buf-flushing
patch, but I wasn't sure he was talking about that patch). So I asked
which code caused the loss of stability to understand if he was talking
about the buf-flushing patch or not, and you replyed saying that you added
wait_on_buffer() in try_to_free_buffers() and the MAP_SHARED problem is
been hided by this trick. I thank you for the reminder about your change
and I apologise if I'm not been clear in my question but I'm still waiting
an answer about what caused the loss of stability. Rik, was you talking
about the buf-flushing patch or about something else? Do you feel that
it's too complicated or messy? Note that I have not proposed it for
inclusion even if I consider more correct than the current code.
Andrea
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Jul 07 2000 - 21:00:09 EST