Bug#593652: grub-common: grub-probe segfault (revised patch)

2010-09-25 Thread Martin Schitter
here is a revised patch. it takes care, that spare (index == 0x) or faulty (index == 0xfffe) drives don't increment array->nr_devs. otherwise the function grub_is_array_readable would return wrong results. sorry, i didn't recognize that in my first attempt. === modified file 'grub-core/di

Bug#593652: grub-common: grub-probe segfault (revisited patch)

2010-09-25 Thread Martin Schitter
here is a revisted patch. it takes care, that spare (index == 0x) or faulty (index == 0xfffe) drives don't increment array->nr_devs. otherwise the function grub_is_array_readable would return wrong results. sorry, i didn't recognize that in my first attempt. === modified file 'grub-core/d

Bug#593652: grub-common: grub-probe segfault

2010-09-25 Thread Martin Schitter
just a few additional remarks about this bug fix in my last post: spare disks within a raid array don't show a useful 'index' number. they may have values like 65535 in their index field. whiteout this fix it's nearly impossible to install grub on any machine with software raid and spare disks

Bug#593652: grub-common: grub-probe segfault

2010-09-25 Thread Martin Schitter
grub2 upstream needs some fixes to accept spare disks in raid arrays. the attached modifications in 'grub-core/drive/raid.c' stopped the segmentation faults on my machine. === modified file 'grub-core/disk/raid.c' --- grub-core/disk/raid.c 2010-09-13 21:59:22 + +++ grub-core/disk/raid.c 2010-