lkml.org 
[lkml]   [1999]   [May]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: All the problems with 2.2.8/2.3.x and bdflush/update
On Fri, 14 May 1999, Zack Weinberg wrote:

>`Being friendly under crash' is an essential behavior which should not
>be made light of.

Sure. I have to admit that if I run a `cvs co linux' and while it's
running I shutdown the machine, then with my patch you have some more fs
corruption, but that's not an issue according to me.

If you stay idle for one minute and then you'll shutdown the machine
without unmount my buffer.c don't lose one bit of data on disk.

If you want to be more conservative (now I use 30 sec of timeout for the
superblock and 60 for data/metadata blocks) you have only to change
/proc/sys/vm/bdflush and my buffer.c will follow you closely (I can agree
also to return to the old timeout even if I would like these new ones
also considering sometime in the future will be fault-tolerant).

>Your jumbo patch tries to do so many different things at once that I
>cannot evaluate it. Could you please separate out the flushtime
>changes versus 2.3.1? Leave everything else alone. Then I will

I am producing a patch against 2.2.8 (I'll upgrade it to 2.2.9 when
patch-linux-kernel will send me the new stable kernel uuencoded ;).

>atime is a red herring. It does affect performance, but mounting

Agreed.

>Are you sure this is safe? Running with debugging compiled in, I saw

I think yes.

>a dirty buffer on the clean list occasionally - just one, and under

You don't need a debugger, you just have SYSRQ+M or SHIFT+something (that
really is racing ;) but I just fixed the race here with a spin_trylock).

>very specific conditions (right at the beginning of bonnie's seek
>tests), so there's another bug somewhere else.

Did you interrupted the kernel in the middle of mark_buffer_dirty?

Andrea Arcangeli


-
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/

\
 
 \ /
  Last update: 2005-03-22 13:51    [W:0.107 / U:23.156 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site