> > Read another 250 times. If one of them succeeds you have your data. > > ...and if it doesn't, it doesn't have the data and cannot perform a > sector remapping.
Relax. That's why I said "in *some* cases". > Why would head tracking fail when a single sector is damaged? It wouldn't. Which is why the the write succeeds successfully, although the data was written to a dubious surface and might not even stay there platter revolution. Ergo, no remapping possible. > doesn't mean that this behaviour is wrong, though. Are you seriously > suggesting that you know better than all of the contributors to the IDE > core in the Linux kernel? If so, go ahead and fix it--send > patches--save the world... everyone will thank you for it! You have an unlimited naive faith in the kernel development process, expecting the outcome to always be the best one can possibly achieve. I know how cr*ppy the kernel can be from my own experience. In fact I'm somewhat surprised that it's used so often in mission-critical applications. Examples: Reading a CD image from a CD hasn't worked properly for any of the 2.4; I haven't checked 2.6 yet as I engage workarounds automatically. SCSI and DMA handling are known to be rock-bottom. I have an old SCSI CD burner which always burnt like clockwork. Now cdrecord has a fit. Why? Neither hardware nor cdrecord have changed. J�rg Schilling told me just now: > >Executing 'mode sense g0' command on Bus 0 Target 5, Lun 0 timeout 40s > >CDB: 1A 00 23 00 06 00 > >cmd finished after 0.014s timeout 40s > >Mode Sense Data 05 A8 00 08 00 00 > >Mode Sense Data 05 A8 00 08 00 00 > >Mode Page Data 00 00 > >./cdrecord: Warning: controller returns zero sized Speed/Dummy information page. > >./cdrecord: Warning: controller returns wrong size for Speed/Dummy information page. > >./cdrecord: Warning: controller returns wrong page 0 for Speed/Dummy information > >page (23). > >./cdrecord: Cannot init drive. > > Looks like a new DMA BUG :-( > > The DMA count is not a multiple of 4 and the first 4 bytes are ok...... How ironic that old but still functional hardware has to be retired because Linux doesn't hack it any more... Wasn't that the argument in favour of Linux? Volker -- Volker Kuhlmann is possibly list0570 with the domain in header http://volker.dnsalias.net/ Please do not CC list postings to me.
