Those 8 failures were due to a single bug.
Here's the fix:

>From c5e82d3b93a3c70e29ae6fd68047502938b12208 Mon Sep 17 00:00:00 2001
From: Jim Meyering <[email protected]>
Date: Tue, 6 Oct 2009 20:10:00 +0200
Subject: [PATCH] gpt: don't malfunction on big-endian systems

Numerous GPT tests would fail when run on e.g., big-endian PPC.
* libparted/labels/gpt.c (gpt_read): Now that we use the
SizeOfPartitionEntry member, be sure to convert from GPT's
little-endian on-disk format to to CPU endianness.
This bug was introduced via commit 14cce9b2, 2009-06-10, "gpt:
fix gpt_read to read all of the partition entries correctly".
* libparted/labels/gpt.c (_header_is_valid): Also convert it here.
Add a test to ensure that the partition entry size is no larger
than the slightly arbitrary UINT32_MAX/16.
* NEWS (Bug fixes): Mention it.
---
 NEWS                   |    3 +++
 libparted/labels/gpt.c |   12 +++++++-----
 2 files changed, 10 insertions(+), 5 deletions(-)

diff --git a/NEWS b/NEWS
index 04741a2..d4147ad 100644
--- a/NEWS
+++ b/NEWS
@@ -18,6 +18,9 @@ GNU parted NEWS                                    -*- 
outline -*-

 ** Bug fixes

+  big-endian systems can once again read GPT partition tables
+  [bug introduced in parted-1.9.0]
+
   ped_partition_is_busy no longer calls libparted's exception handler,
   since doing so caused trouble with anaconda/pyparted when operating on
   dmraid devices.
diff --git a/libparted/labels/gpt.c b/libparted/labels/gpt.c
index cc9bcdc..b4549ef 100644
--- a/libparted/labels/gpt.c
+++ b/libparted/labels/gpt.c
@@ -608,8 +608,10 @@ _header_is_valid (const PedDevice* dev, 
GuidPartitionTableHeader_t* gpt)
         * the SizeOfPartitionEntry must be a multiple of 8 and
         * no smaller than the size of the PartitionEntry structure.
         */
-       uint32_t sope = gpt->SizeOfPartitionEntry;
-       if (sope % 8 != 0 || sope < sizeof(GuidPartitionEntry_t) )
+       uint32_t sope = PED_LE32_TO_CPU (gpt->SizeOfPartitionEntry);
+       if (sope % 8 != 0
+            || sope < sizeof (GuidPartitionEntry_t)
+            || (UINT32_MAX >> 4) < sope)
                return 0;

        origcrc = gpt->HeaderCRC32;
@@ -911,7 +913,8 @@ gpt_read (PedDisk * disk)
        if (!_parse_header (disk, gpt, &write_back))
                goto error_free_gpt;

-       ptes_sectors = ped_div_round_up (gpt->SizeOfPartitionEntry
+       uint32_t p_ent_size = PED_LE32_TO_CPU (gpt->SizeOfPartitionEntry);
+       ptes_sectors = ped_div_round_up (p_ent_size
                                         * gpt_disk_data->entry_count,
                                         disk->dev->sector_size);

@@ -926,8 +929,7 @@ gpt_read (PedDisk * disk)

        for (i = 0; i < gpt_disk_data->entry_count; i++) {
                GuidPartitionEntry_t* pte
-                 = (GuidPartitionEntry_t*) ((char *)ptes + i
-                                            * gpt->SizeOfPartitionEntry);
+                 = (GuidPartitionEntry_t*) ((char *)ptes + i * p_ent_size);
                PedPartition* part;
                PedConstraint* constraint_exact;

--
1.6.5.rc2.204.g8ea19

_______________________________________________
parted-devel mailing list
[email protected]
http://lists.alioth.debian.org/mailman/listinfo/parted-devel

Reply via email to