| To: | Wessel Dankers <wsl@xxxxxxxxxxxx> |
|---|---|
| Subject: | Re: The mmap() problem is back |
| From: | Steve Lord <lord@xxxxxxx> |
| Date: | 05 Aug 2002 17:02:03 -0500 |
| Cc: | linux-xfs@xxxxxxxxxxx |
| In-reply-to: | <20020805184232.GD863@xxxxxxxxxxxx> |
| References: | <20020805010731.F24705@xxxxxxxxxxxxxxx> <3674.1028557285@xxxxxxxxxxxxxxxxxxxxx> <20020805184232.GD863@xxxxxxxxxxxx> |
| Sender: | owner-linux-xfs@xxxxxxxxxxx |
Can you try this change and let me know if it fixes it for you,
I did not check this in because I do not want to fix it this way,
but it will confirm where the problem is coming from:
Steve
===========================================================================
Index: linux/fs/xfs/linux/xfs_iops.c
===========================================================================
--- /usr/tmp/TmpDir.27400-0/linux/fs/xfs/linux/xfs_iops.c_1.167 Mon Aug 5
17:02:15 2002
+++ linux/fs/xfs/linux/xfs_iops.c Mon Aug 5 16:38:56 2002
@@ -559,9 +559,11 @@
/* If we are doing writes at the end of the file,
* allocate in chunks
*/
+/****
if (create && (offset >= inode->i_size) && !(flags & PBF_SYNC))
size = 1 << XFS_WRITE_IO_LOG;
else
+***/
size = 1 << inode->i_blkbits;
VOP_BMAP(vp, offset, size,
--
Steve Lord voice: +1-651-683-3511
Principal Engineer, Filesystem Software email: lord@xxxxxxx
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: The mmap() problem is back, Steve Lord |
|---|---|
| Next by Date: | Re: HELP - serious problem, weird filesystem behaviour ..., Tim Shimmin |
| Previous by Thread: | Re: The mmap() problem is back, Steve Lord |
| Next by Thread: | Re: The mmap() problem is back, Wessel Dankers |
| Indexes: | [Date] [Thread] [Top] [All Lists] |