| Issue 34: | crash when running OHSM realloc module for the first time on a mount point after mkfs.ext2 on it. | |
| 1 person starred this issue and may be notified of changes. | Back to list |
|
What steps will reproduce the problem? 1. Create a LV and do a mkfs.ext2 on that. 2. mount it. 3. insmod relloc module What is the expected output? What do you see instead? It should work well. It is seen that on the very first run it happens always. this is reproducible. But If we insmod for the second time and later on .... It works perfectly fine. Looking into the issue. Also, you have to reboot after the crash to be able to insmod the realloc module.
Jan 14, 2009
Project Member
#1
sandeepksinha
Status:
In-progress
Jan 14, 2009
Yes Sir, The issue is resolved. The problem that we were facing was that during relocation when we were creating a temporary or ghost inode, we were not setting the i_data fields of the inode to zero because of which it was spanning to a block number which doesn't exist and while reading it's buffer it was giving kernel paging request error. Second issue was faced after deleting the temporary or ghost inode. The ghost inode was marked dirty and later we deleted the ghost inode and when the sync takes place after relocation got over, the ghost inode doesn't exist hence this was the potential BUG. well all issue are now taken and solved. The code is of relocation is now working fine.
Status:
Done
Cc: fscops checkout.vineet sneha.hendre sandeepksinha bharati.alatgi imreckless postrishi rohitvashist2kk3 |
||||||||||