summaryrefslogtreecommitdiffstats
path: root/meta/recipes-devtools/rpm/rpm/debugedit-segv.patch
diff options
context:
space:
mode:
Diffstat (limited to 'meta/recipes-devtools/rpm/rpm/debugedit-segv.patch')
-rw-r--r--meta/recipes-devtools/rpm/rpm/debugedit-segv.patch35
1 files changed, 35 insertions, 0 deletions
diff --git a/meta/recipes-devtools/rpm/rpm/debugedit-segv.patch b/meta/recipes-devtools/rpm/rpm/debugedit-segv.patch
new file mode 100644
index 0000000000..bd9169381f
--- /dev/null
+++ b/meta/recipes-devtools/rpm/rpm/debugedit-segv.patch
@@ -0,0 +1,35 @@
1There are cases, especially on PPC and MIPS, where the data address
2returned is 0, but the size is not 0.
3
4It appears to happen when the sections headers are similar to:
5
6 [21] .data PROGBITS 000239c0 0139c0 000010 00 WA 0 0 8
7 [22] .got PROGBITS 000239d0 0139d0 000014 04 WAX 0 0 4
8 [23] .plt NOBITS 000239e4 0139e4 000234 00 WAX 0 0 4
9 [24] .bss NOBITS 00023c18 0139e4 0001c8 00 WA 0 0 8
10 [25] .comment PROGBITS 00000000 0139e4 000011 01 MS 0 0 1
11 [26] .debug_aranges PROGBITS 00000000 0139f8 000d68 00 0 0 8
12
13Sections 23 and 24 (.plt and .bss) which are NOBITS have a loaded data address
14of 0, but a size != 0.
15
16This could be a bug in libelf...
17
18Upstream-status: Pending
19
20Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
21
22Index: rpm-5.4.9/tools/debugedit.c
23===================================================================
24--- rpm-5.4.9.orig/tools/debugedit.c
25+++ rpm-5.4.9/tools/debugedit.c
26@@ -1434,7 +1434,8 @@ handle_build_id (DSO *dso, Elf_Data *bui
27 auto inline void process (const void *data, size_t size)
28 {
29 memchunk chunk = { .data = (void *) data, .size = size };
30- hashFunctionContextUpdateMC (&ctx, &chunk);
31+ if (data != NULL && size != 0)
32+ hashFunctionContextUpdateMC (&ctx, &chunk);
33 }
34 union
35 {