[xfs-masters] [Bug 799] New: xfsdump fails for large page sizes

[bugzilla-daemon at oss.sgi.com]

http://oss.sgi.com/bugzilla/show_bug.cgi?id=799

           Summary: xfsdump fails for large page sizes
           Product: Linux XFS
           Version: unspecified
          Platform: Other
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: xfsdump
        AssignedTo: xfs-masters at oss.sgi.com
        ReportedBy: sandeen-xfs at sandeen.net
                CC: wkendall at sgi.com


On ppc64 with 64k pages, xfsdump fails with several assertions about page size
in relation to various other sizes.  In particular, these:

--- xfsdump-2.2.48.orig/restore/inomap.c
+++ xfsdump-2.2.48/restore/inomap.c
@@ -197,8 +197,8 @@ inomap_restore_pers( drive_t *drivep,
         */
        ASSERT( INOPERSEG == ( sizeof( (( seg_t * )0 )->lobits ) * NBBY ));
        ASSERT( sizeof( hnk_t ) == HNKSZ );
-       ASSERT( HNKSZ >= pgsz );
-       ASSERT( ! ( HNKSZ % pgsz ));
+       //ASSERT( HNKSZ >= pgsz );
+       //ASSERT( ! ( HNKSZ % pgsz ));
        ASSERT( sizeof( pers_t ) <= PERSSZ );
 
        /* get inomap info from media hdr
@@ -224,8 +224,8 @@ inomap_restore_pers( drive_t *drivep,
 
        /* mmap the persistent hdr and space for the map
         */
-       ASSERT( sizeof( hnk_t ) * ( size_t )hnkcnt >= pgsz );
-       ASSERT( ! ( sizeof( hnk_t ) * ( size_t )hnkcnt % pgsz ));
+       //ASSERT( sizeof( hnk_t ) * ( size_t )hnkcnt >= pgsz );
+       //ASSERT( ! ( sizeof( hnk_t ) * ( size_t )hnkcnt % pgsz ));
        persp = ( pers_t * ) mmap_autogrow(
                                     PERSSZ
                                     +
@@ -355,8 +355,8 @@ inomap_sync_pers( char *hkdir )
        /* sanity checks
         */
        ASSERT( sizeof( hnk_t ) == HNKSZ );
-       ASSERT( HNKSZ >= pgsz );
-       ASSERT( ! ( HNKSZ % pgsz ));
+       //ASSERT( HNKSZ >= pgsz );
+       //ASSERT( ! ( HNKSZ % pgsz ));
 
        /* only needed once per session
         */

all trip.  xfsdump seems a bit confused about "PGSZ" (hard-coded to 4096) and
pgsz (actual machine page size) and what is part of the dump format, what is
optimal IO, what is required for proper IO alignment, etc.  However, I'll note
that simply commenting out the above ASSERTs on 64k ppc64 did allow all of the
xfsdump xfsqa tests to pass (with a kernel that has my fixed compat ioctls
hooked up, that is).

-- 
Configure bugmail: http://oss.sgi.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.