Hi Thomas,

never tried a local copy with 1 or 2G files.

This bug renders my system useless as I move big files around, for that
reason I now moved to FreeBSD.

What did you do about this issue?? Installed older version of Ubuntu?

Greetz,

Huib

> Have you tried copying locally with very large files (1-2 GB)?  The
> corruption didn't occur for me with smaller files locally.
>
> On Sun, Aug 17, 2008 at 8:15 AM, Huib Smeets <[EMAIL PROTECTED]> wrote:
>> Hi,
>>
>> on a clean install of ubuntu 8.04 (AMD kernel version) I get this very
>> same error: random bytes get mangled but the byte value as documented
>> above always differ value 10!
>>
>> After reverting back to the i386 kernel I still have the same error. The
>> corruption only happens in network traffic (ftp or samba) when I upload
>> the file to the linux server.
>>
>> I changed from AMD to I386 kernel --> problem persists
>> I swapped NIC's from onboard attansic/atheros (ATL1) to Intel Pro
>> (E1000E) --> problem persists
>> copy files locally -> no problem
>> I installed FreeBSD 7 -> no problem
>>
>> Seems to be network related but not the specific driver and it happens
>> only to the bigger files (>20Mbytes)
>>
>> Hardware:
>> - Asus M3A-H/HDMI 780G
>> - AMD Athlon 4450e
>>
>> --
>> Regression: File corruption on amd64/nVidia with 2.6.22-14-generic
>> https://bugs.launchpad.net/bugs/162478
>> You received this bug notification because you are a direct subscriber
>> of the bug.
>>
>
>
> --
> http://resc.smugmug.com/
> AIM: ferret taskforce
>
> --
> Regression: File corruption on amd64/nVidia with 2.6.22-14-generic
> https://bugs.launchpad.net/bugs/162478
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in “linux-ubuntu-modules-2.6.22” source package in Ubuntu:
> Invalid
>
> Bug description:
> I upgraded from Feisty to Gutsy using the update manager, and a few days
> later I started having problems.  First, it showed up as bad downloads
> using apt-get (one of which happened to be libc6, which made my system
> unbootable for a while).  Then apt-get started segfaulting (bad files in
> /var/cache/apt).  Later, I downloaded a 20MB zip file I was trying to use
> 3 times, and got 3 different files.
>
> At first, I thought it was a network problem, but now I have reproduced it
> using only disk accesses.  I create a 1GB file from /dev/urandom, and copy
> it to several other files, to force all of RAM to be used (I have 2GB).
> Then, I run md5sum on all of the files and get different results for each
> one.  The differences in the files are really suggestive of a memory
> problem, but I have run memtest86+ for a couple of passes and it finds
> nothing.
>
> Output of cmp -lb original.bin copied1.bin:
>  584294856 227 M-^W 237 M-^_
>  593924424  64 4     74 <
>  633975304 224 M-^T 234 M-^\
>  640479752 202 M-^B 212 M-^J
>  644645384 163 s    173 {
>  656228872 322 M-R  332 M-Z
>  662528520  46 &     56 .
>  662782472 204 M-^D 214 M-^L
>  676360712 167 w    177 ^?
>  782352584 207 M-^G 217 M-^O
> Bit 4 is always switched from 0 to 1.
>
> I don't think that this is a hardware problem, because I have run
> memtest86+ for a couple of passes, and it found no problems.  It also does
> not occur on 2.6.20-16, as far as I can tell (I've tried the procedure
> above several times and got no differences in the files).  I also tried
> booting with iommu=soft, which did not fix the problem.
>
> I think that this is a file writing problem, rather than a reading
> problem, because I have checked the md5sums of the corrupted files and
> they are consistent between kernel versions/reboots.
>
> I'll attach dmesg, cpuinfo, lspci, etc. in a moment.
>

-- 
Regression: File corruption on amd64/nVidia with 2.6.22-14-generic
https://bugs.launchpad.net/bugs/162478
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to