rpms/kernel/devel squashfs-fix-page-aligned-data.patch, NONE, 1.1 kernel.spec, 1.1408, 1.1409

[Kyle McMartin]

Author: kyle

Update of /cvs/pkgs/rpms/kernel/devel
In directory cvs1.fedora.phx.redhat.com:/tmp/cvs-serv31906

Modified Files:
	kernel.spec 
Added Files:
	squashfs-fix-page-aligned-data.patch 
Log Message:
* Tue Mar 10 2009 Kyle McMartin <kyle at redhat.com> 2.6.29-0.229.rc7.git3
- squashfs-fix-page-aligned-data.patch:
   Theoretically valid images are hitting the if (page == pages) check,
   which is suspect happens when the data is page aligned... We should
   probably be checking for page+1.


squashfs-fix-page-aligned-data.patch:

--- NEW FILE squashfs-fix-page-aligned-data.patch ---
diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c
index 321728f..2b7598e 100644
--- a/fs/squashfs/block.c
+++ b/fs/squashfs/block.c
@@ -185,7 +185,7 @@ int squashfs_read_data(struct super_block *sb, void **buffer, u64 index,
 			}
 
 			if (msblk->stream.avail_out == 0) {
-				if (page == pages) {
+				if (page > pages) {
 					ERROR("zlib_inflate tried to "
 						"decompress too much data, "
 						"expected %d bytes.  Zlib "


Index: kernel.spec
===================================================================
RCS file: /cvs/pkgs/rpms/kernel/devel/kernel.spec,v
retrieving revision 1.1408
retrieving revision 1.1409
diff -u -r1.1408 -r1.1409
--- kernel.spec	10 Mar 2009 17:10:09 -0000	1.1408
+++ kernel.spec	11 Mar 2009 03:24:59 -0000	1.1409
@@ -688,6 +688,8 @@
 
 Patch9002: cpufreq-add-atom-to-p4-clockmod.patch
 
+Patch9003: squashfs-fix-page-aligned-data.patch
+
 %endif
 
 BuildRoot: %{_tmppath}/kernel-%{KVERREL}-root
@@ -1232,6 +1234,8 @@
 
 ApplyPatch cpufreq-add-atom-to-p4-clockmod.patch
 
+ApplyPatch squashfs-fix-page-aligned-data.patch
+
 # END OF PATCH APPLICATIONS
 
 %endif
@@ -1815,6 +1819,12 @@
 # and build.
 
 %changelog
+* Tue Mar 10 2009 Kyle McMartin <kyle at redhat.com> 2.6.29-0.229.rc7.git3
+- squashfs-fix-page-aligned-data.patch:
+   Theoretically valid images are hitting the if (page == pages) check,
+   which is suspect happens when the data is page aligned... We should
+   probably be checking for page+1.
+
 * Tue Mar 10 2009 Tom "spot" Callaway <tcallawa at redhat.com>
 - disable CONFIG_AGP on sparc64