xfs_fsr: more selinux fixes

Commit:

1adfe5c xfs_fsr: fix SWAPEXT failures under selinux

attempted to fix up the fork offset under selinux, where the temp file is created with a local attribute, but the target file has remote attributes; this can lead to a smaller data area in the temp inode, without enough room to swap extents from the target inode. I remedied this by pushing the temp file attribute to remote, but *only* if the target file's attr was also remote.

However, I have a case from the field where the parent dir and the target file both have a context of:

system_u:object_r:samba_share_t:s0

but new files created in the dir have a context of

unconfined_u:object_r:samba_share_t:s0

This means the temp file has a smaller forkoff, and less space in the inode for data, so we fail to swap the extents between the two, because they don't fit.

The following patch fixes this by allowing xfs_fsr to kick the tempfile's attr out of local format even if the target file's attr is local, if this will move the forkoff in the right direction. This does pass all our fsr xfstests, though I'm not sure we have any real coverage of fsr under selinux...

The only functional change is the test at the very end of the patch; the rest is comments, ascii art, and removing the now-extraneous XFS_IOC_FSGETXATTRA ioctl.

Signed-off-by Eric Sandeen <[email protected]>

---

fsr/xfs_fsr.c +0 -0

index c8ef18f..68b9819 100644

fsr/xfs_fsr.c +0 -4

 			/*
-			 * If target attr area is less than the temp's (diff < 0)
-			 * and the target is not local, write a big attr to
-			 * the temp file to knock the attr out of local format,
-			 * to match the target.  (This should actually *increase*

fsr/xfs_fsr.c +0 -0

 				char val[2048];
 				memset(val, 'X', 2048);
 				if (fsetxattr(tfd, name, val, 2048, 0)) {

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

Quoted 42 lines · Eric Sandeen

Commit:

1adfe5c xfs_fsr: fix SWAPEXT failures under selinux

attempted to fix up the fork offset under selinux, where the temp file is created with a local attribute, but the target file has remote attributes; this can lead to a smaller data area in the temp inode, without enough room to swap extents from the target inode. I remedied this by pushing the temp file attribute to remote, but *only* if the target file's attr was also remote.

However, I have a case from the field where the parent dir and the target file both have a context of:

system_u:object_r:samba_share_t:s0

but new files created in the dir have a context of

unconfined_u:object_r:samba_share_t:s0

This means the temp file has a smaller forkoff, and less space in the inode for data, so we fail to swap the extents between the two, because they don't fit.

The following patch fixes this by allowing xfs_fsr to kick the tempfile's attr out of local format even if the target file's attr is local, if this will move the forkoff in the right direction. This does pass all our fsr xfstests, though I'm not sure we have any real coverage of fsr under selinux...

The only functional change is the test at the very end of the patch; the rest is comments, ascii art, and removing the now-extraneous XFS_IOC_FSGETXATTRA ioctl.

Signed-off-by Eric Sandeen <[email protected]>

---

fsr/xfs_fsr.c +0 -0

index c8ef18f..68b9819 100644

...

Quoted 11 lines

fsr/xfs_fsr.c +4 -6

 			/*
-			 * If target attr area is less than the temp's (diff < 0)
-			 * and the target is not local, write a big attr to
-			 * the temp file to knock the attr out of local format,
-			 * to match the target.  (This should actually *increase*
-			 * the temp file's forkoffset when the attr moves out
-			 * of the inode)
+			 * If target attr area is less than the temp's
+			 * (diff < 0), write a big attr to the temp file to knock
+			 * the attr out of local format.
+			 * (This should actually *increase* the temp file's

fsr/xfs_fsr.c fsr_setup_attr_fork( · line 1133

Quoted 4 lines

fsr/xfs_fsr.c +2 -1

+			 * forkoffset when the attr moves out of the inode)
 			 */
- 			if (diff < 0 && fsx.fsx_nextents > 0) {
+ 			if (diff < 0) {

Space before tab issue on the line above. Looks fine otherwise, but I wonder if it would also be a good idea to add an informative fsrprintf() here if we proceed when (fsx.fsx_nextents > 0)?

Commit:

1adfe5c xfs_fsr: fix SWAPEXT failures under selinux

attempted to fix up the fork offset under selinux, where the temp file is created with a local attribute, but the target file has remote attributes; this can lead to a smaller data area in the temp inode, without enough room to swap extents from the target inode. I remedied this by pushing the temp file attribute to remote, but *only* if the target file's attr was also remote.

However, I have a case from the field where the parent dir and the target file both have a context of:

system_u:object_r:samba_share_t:s0

but new files created in the dir have a context of

unconfined_u:object_r:samba_share_t:s0

This means the temp file has a smaller forkoff, and less space in the inode for data, so we fail to swap the extents between the two, because they don't fit.

The following patch fixes this by allowing xfs_fsr to kick the tempfile's attr out of local format even if the target file's attr is local, if this will move the forkoff in the right direction. This does pass all our fsr xfstests, though I'm not sure we have any real coverage of fsr under selinux...

The only functional change is the test at the very end of the patch; the rest is comments, ascii art, and removing the now-extraneous XFS_IOC_FSGETXATTRA ioctl.

Signed-off-by Eric Sandeen <[email protected]>

---

V2: Fix a whitespace issue.

Brian, I didn't add the debug printf; we don't do it for any other attr operations unless they fail, and I don't think it's needed here either. And I couldn't fit it gracefully into 80 cols. ;)

fsr/xfs_fsr.c +8 -20

index b902acc..2bae1d0 100644
@@ -1115,14 +1115,12 @@ fsr_setup_attr_fork(
 			}
 			continue;
 		} else if (i == 0) {
-			struct fsxattr	fsx;
 			/*
 			 * First pass, and temp file already has an inline
 			 * xattr, probably due to selinux.
 			 *
 			 * It's *possible* that the temp file attr area
-			 * is larger than the target file's, if the 
-			 * target file's attrs are not inline:
+			 * is larger than the target file's:
 			 *
 			 *  Target		 Temp
 			 * +-------+ 0		+-------+ 0

@@ -1132,28 +1130,18 @@ fsr_setup_attr_fork(
 			 * |	   |		v-------v forkoff
 			 * |	   |		|       |
 			 * v-------v forkoff	| Attr  | local
-			 * | Attr  | ext/btree	|       |
+			 * | Attr  | 		|       |
 			 * +-------+		+-------+
-			 *
-			 * FSGETXATTRA will tell us nr of attr extents in
-			 * target, if any.  If none, it's local:
 			 */
 
-			memset(&fsx, 0, sizeof(fsx));
-			if (ioctl(fd, XFS_IOC_FSGETXATTRA, &fsx)) {
-				fsrprintf(_("FSGETXATTRA failed on target\n"));
-				return -1;
-			}
-
 			/*
-			 * If target attr area is less than the temp's (diff < 0)
-			 * and the target is not local, write a big attr to
-			 * the temp file to knock the attr out of local format,
-			 * to match the target.  (This should actually *increase*
-			 * the temp file's forkoffset when the attr moves out
-			 * of the inode)
+			 * If target attr area is less than the temp's
+			 * (diff < 0) write a big attr to the temp file to knock
+			 * the attr out of local format.
+			 * (This should actually *increase* the temp file's
+			 * forkoffset when the attr moves out of the inode)
 			 */
- 			if (diff < 0 && fsx.fsx_nextents > 0) {
+			if (diff < 0) {
 				char val[2048];
 				memset(val, 'X', 2048);
 				if (fsetxattr(tfd, name, val, 2048, 0)) {

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs