Am 10.08.21 um 08:16 schrieb Carlo Marcelo Arenas Belón:
Thanks,
in the discussion above René[1] proposed a fix for UBsan issues that were
reported and that it is still missing.
my version of it didn't require the extra 4 bytes or showed issues with
notes so is probably incomplete and should be replaced from the original
if possible, but follows below:
With your three patches plus the one below t3301-notes.sh and several more
fail on an Apple M1. Adding an unused int member to struct leaf_node fixes
that. I didn't dig deeper into the notes code to understand the actual
issue, though.
quoted hunk ↗ jump to hunk
Carlo
[1] https://lore.kernel.org/git/bab9f889-ee2e-d3c3-0319-e297b59261a0@web.de/ (local)
+CC René for advise
Date: Sun, 8 Aug 2021 20:45:56 -0700
Subject: [PATCH] build: fixes for SANITIZE=undefined (WIP)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
mostly from instructions/code provided by René in :
https://lore.kernel.org/git/20210807224957.GA5068@dcvr/ (local)
tested with Xcode in macOS 11.5.1 (x86_64)
---
hash.h | 2 +-
object-file.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/hash.h b/hash.h
index 27a180248f..3127ba1ef8 100644
--- a/hash.h
+++ b/hash.h
@@ -115,7 +115,7 @@ static inline void git_SHA256_Clone(git_SHA256_CTX *dst, const git_SHA256_CTX *s
struct object_id {
unsigned char hash[GIT_MAX_RAWSZ];
- int algo;
+ uint8_t algo;
};
/* A suitably aligned type for stack allocations of hash contexts. */diff --git a/object-file.c b/object-file.c
index 374f3c26bf..2fa282a9b4 100644
--- a/object-file.c
+++ b/object-file.c
@@ -2406,7 +2406,7 @@ struct oidtree *odb_loose_cache(struct object_directory *odb,
struct strbuf buf = STRBUF_INIT;
size_t word_bits = bitsizeof(odb->loose_objects_subdir_seen[0]);
size_t word_index = subdir_nr / word_bits;
- size_t mask = 1 << (subdir_nr % word_bits);
+ size_t mask = 1U << (subdir_nr % word_bits);
uint32_t *bitmap;
if (subdir_nr < 0 ||
The first hunk is about alignment (and missing the notes fix, as mentioned).
The second hunk is about shifting a signed 32-bit value 31 places to the
left, which is undefined (because technically there are only 31 value bits).
Those are different issues and they should be addressed by separate patches,
I think. That's why I submitted a patch for the the second one in
http://public-inbox.org/git/bab9f889-ee2e-d3c3-0319-e297b59261a0@web.de/.
René