On Saturday 13 November 2004 09:02, Jim Cornette wrote:
Erik Hemdal wrote:
>>>>md5sum -c MD5SUM
>>>>I've already indicated this once. After successful MD5SUM check
>>>> - boot with 'linux ide=nodma' and then do the MediaCheck.)
> I think this comment was from Satish about ide=nodma. Why is that
> necessary? I've encountered the mediacheck failures similarly for
> the first time ever. If the drive operates normally, why do we
> need nodma in order to check only disk2 and disk3 of the CD set?
> I'm not intending this as disrespectful to anyone; I'm truly
> curious about it. Out of the four install images, all four pass
> the MD5 checksum test, and two out of four pass the mediacheck.
> On my CD writer, cdrecord reports that the buffer was always
> adequately filled and burnfree protection was never needed. So
> assuming the integrity of the media is OK (seems to be, since
> cdrecord reports no errors and the CD is readable), I'd suspect a
> problem with the original ISO image. Whether you use DMA or not on
> your drive doesn't seem to have any bearing on things....else
> other I/O would have trouble too.
> Am I missing something?
Check the fedora-test-list for discussions regarding the mediacheck
failure. Padding the discs with at least 150 kb was suggested,
loading the ide-scsi or using the 2.4 kernel was suggested. I guess
there is some breakage which happened with the change to the 2.6
I'm glad that adding ide=nodma to the installer before a mediacheck
works to verify the discs successfully. This also probably helps
ensure more reliable installs with systems that show the mediacheck
false result during verifying.
I feared a mass explosion of coasters because of this problem and no
reference to the problem within the release notes. For those reading
this list, the coaster collection is probably lower.
For installers not on the list and attempting an install, the
mirrors, reburning discs multiple times and giving up on the
release might be at higher levels.
After having a hell of a time getting them, I burnt all 9 disks
yesterday usng k3b from a konstuct built kde3.3 install, makeing k3b
verify the burns, using kernel 2.6.10-rc1-bk20. dma is enabled and
the minimum buffer fill during any of the burns at about 32x was 98%.
Perhaps something has sneaked into the distribution kernel that
doesn't need to be there?
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.28% setiathome rank, not too shabby for a WV hillbilly
attorneys please note, additions to this message
by Gene Heskett are:
Copyright 2004 by Maurice Eugene Heskett, all rights reserved.