Pretty much the SUBJ.<br />
<br />
On a system I'm administer I've got a fresh installation of CentOS 6.2 x86-64. There are a lot of HDDs attached to that box (10x1.5Tb SATA) and it's pretty common for modern hard drives to hit so-called "non-permanent bad sectors", which can't be read by HDD, but can be fully recovered when written with zeroes using low-level ATA command. A way I used to do this on my home PC (running LFS-like system based on Fedora 14 sources) was to issue something like:<br />
<br />
# hdparm --yes-i-know-what-i-am-doing --write-sector 722431704 /dev/sdc<br />
<br />
Today I've run up into segmentation fault issue trying to execute the above command on the CentOS 6 installation. Same SIGSEGV also happen when trying to use "--read-sector" command. Upgrading hdparm package to the one from Fedora 13 (namely hdparm-9.27-1.fc13.x86_64.rpm) fixes the issue.<br />
<br />
I wanted to try to debug the issue myself but the problem is that debuginfo CentOS 6 repo lacks debugging symbols for hdparm package and I haven't got enough time currently to manually rebuild hdparm using mock in order to produce the required debugging symbols myself.
↧
