Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 12 Nov 2006 15:43:25 +0100
From:      Andre Oppermann <andre@freebsd.org>
To:        Pawel Worach <pawel.worach@gmail.com>
Cc:        freebsd-current@freebsd.org, amistry@am-productions.biz
Subject:   Re: sendfile data corruption
Message-ID:  <4557330D.3010009@freebsd.org>
In-Reply-To: <4555BA65.4020603@gmail.com>
References:  <200611021653.kA2GrRWx008044@repoman.freebsd.org>	<455530E0.5090000@gmail.com> <4555B45D.6020800@freebsd.org> <4555BA65.4020603@gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Pawel Worach wrote:
> Andre Oppermann wrote:
> 
>>
>> I'm looking into the problem.  Please try a binary FTP transfer as well
>> and check if the checksums match. ftpd uses sendfile(2) as well but w/o
>> headers or trailers and does the send in one swoop.
>>
> 
> Oh, didn't think of that, ftpd is ok, transferring a 64MB file does not 
> trash it. Meanwhile a couple of other things where tested, SMP disabled 
> (removed from kernel config), added some printf's which when printing to 
> a serial console moves the offset where the breakage begins to 
> 0x01000000, sometimes.

OK, I found the bug.  The sent byte count reporting was incorrect.  While
doing the sendfile(2) rewrite I got lost in the mixup of the FreeBSD 4.x
bug for bug compatibility.

Please try this patch:

  http://people.freebsd.org/~andre/sendfile_fix-20061112.diff

It fixes apache 2.0.59 for me.  For some reason lighttpd didn't suffer
from this problem, even w/o the fix.  Unfortunately that's what I tested
the rewrite against.

-- 
Andre



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4557330D.3010009>