This thread contains a patchset. You're looking at the original emails, but you may wish to use the patch review UI. Review patch

[PATCH] build_target: Fix possible out-of-bounds egde case

Message ID
DKIM signature
Download raw message
Patch: +1 -5
Detected on Tinderbox with FORTIFY_SOURCE=3.

Bug: https://bugs.gentoo.org/882295
Hey there,

The Tinderbox spotted a potential error in determine_target_build_name while
running tests.  I traced it back to a bad concat, fix included.

Tested on x86_64-pc-linux-gnu, though I'm unsure if I got all the
configurations or if the contract of the function is still fulfilled following
this change (but it appears to be to me).

Thanks in advance, and have a great night!

 src/functions/kernel/build_target.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/src/functions/kernel/build_target.c b/src/functions/kernel/build_target.c
index 07b20e29..2ecc27c4 100644
--- a/src/functions/kernel/build_target.c
+++ b/src/functions/kernel/build_target.c
@@ -307,7 +307,6 @@ determine_target_build_name(struct workspace *wk, struct obj_build_target *tgt,
	obj name_pre, obj name_suff, char plain_name[BUF_SIZE_2k])
	const char *pref, *suff, *ver_suff = NULL;
	uint32_t i;

	switch (tgt->type) {
	case tgt_executable:
@@ -341,10 +340,7 @@ determine_target_build_name(struct workspace *wk, struct obj_build_target *tgt,
		suff = get_cstr(wk, name_suff);

	i = snprintf(plain_name, BUF_SIZE_2k, "%s%s", pref, get_cstr(wk, tgt->name));
	if (suff) {
		snprintf(&plain_name[i], BUF_SIZE_2k, ".%s", suff);
	snprintf(plain_name, BUF_SIZE_2k, "%s%s%s%s", pref, get_cstr(wk, tgt->name), suff ? "." : "", suff ? suff : "");

	tgt->build_name = make_strf(wk, "%s%s%s", plain_name, ver_suff ? "." : "", ver_suff ? ver_suff : "");
	return true;
Reply to thread Export thread (mbox)