]> git.ipfire.org Git - thirdparty/openembedded/openembedded-core.git/commit
dmidecode: fixup for CVE-2023-30630
authorSean Nyekjaer <sean@geanix.com>
Mon, 16 Oct 2023 12:21:58 +0000 (14:21 +0200)
committerSteve Sakoman <steve@sakoman.com>
Mon, 16 Oct 2023 16:04:57 +0000 (06:04 -1000)
commit0bc69dc078c39381a39789d3c5fff673d7da994c
tree33d9941a63c77ce50decbfa8e120bbc0166b1a11
parenta1138dfc1e9394966dcca8f6259767a8ce5ca4cb
dmidecode: fixup for CVE-2023-30630

The previous CVE-2023-30630_1.patch picked only the patch
"dmidecode: Write the whole dump file at once" d8cfbc808f.
But there was a refactoring which does not allow to cherry-pick it fast
forward. Resolving this conflict was not correctly done. The patch was:

+    u32 len;
+    u8 *table;
...
-    if (!(opt.flags & FLAG_QUIET))
-        pr_comment("Writing %d bytes to %s.", crafted[0x05],
-                   opt.dumpfile);
-    write_dump(0, crafted[0x05], crafted, opt.dumpfile, 1);
+    dmi_table_dump(crafted, crafted[0x05], table, len);

It looks like the variables len and table have been added without
initialization.
Now this problem is solved by applying the previous refactoring as
well. Patch 1 gets replaced by Patch 1a and Patch 1b. Patch 2..4 are
rebased without changes.

This is basically the same patch as in kirkstone:
ea069a94a2 dmidecode: fixup for CVE-2023-30630

Signed-off-by: Sean Nyekjaer <sean@geanix.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
meta/recipes-devtools/dmidecode/dmidecode/CVE-2023-30630_1a.patch [new file with mode: 0644]
meta/recipes-devtools/dmidecode/dmidecode/CVE-2023-30630_1b.patch [new file with mode: 0644]
meta/recipes-devtools/dmidecode/dmidecode/CVE-2023-30630_2.patch
meta/recipes-devtools/dmidecode/dmidecode/CVE-2023-30630_3.patch
meta/recipes-devtools/dmidecode/dmidecode/CVE-2023-30630_4.patch
meta/recipes-devtools/dmidecode/dmidecode_3.4.bb