Re: [PATCH] more oideq/hasheq conversions

2018-10-02 Thread Jacob Keller
On Tue, Oct 2, 2018 at 2:19 PM Jeff King  wrote:
>
> We added faster equality-comparison functions for hashes in
> 14438c4497 (introduce hasheq() and oideq(), 2018-08-28). A
> few topics were in-flight at the time, and can now be
> converted. This covers all spots found by "make coccicheck"
> in master (the coccicheck results were tweaked by hand for
> style).
>
> Signed-off-by: Jeff King 
> ---
> Jake: I was surprised that this was not a "patch 2" on top of your
> earlier coccicheck patch. Apologies if you were planning to send it out.
>

Nope, I hadn't gotten around to this yet, thanks!

The conversions also look good to me :)

Regards,
Jake


Re: [PATCH] more oideq/hasheq conversions

2018-10-02 Thread Derrick Stolee

On 10/2/2018 5:19 PM, Jeff King wrote:

We added faster equality-comparison functions for hashes in
14438c4497 (introduce hasheq() and oideq(), 2018-08-28). A
few topics were in-flight at the time, and can now be
converted. This covers all spots found by "make coccicheck"
in master (the coccicheck results were tweaked by hand for
style).

Signed-off-by: Jeff King 
---
Jake: I was surprised that this was not a "patch 2" on top of your
earlier coccicheck patch. Apologies if you were planning to send it out.

This doesn't conflict with anything in "pu", so it's a reasonable time
to apply it. There are a few lingering cases in pu, so another option is
to wait a week or two and see if they get merged.

These conversions look good to me!

Reviewed-by: Derrick Stolee 


[PATCH] more oideq/hasheq conversions

2018-10-02 Thread Jeff King
We added faster equality-comparison functions for hashes in
14438c4497 (introduce hasheq() and oideq(), 2018-08-28). A
few topics were in-flight at the time, and can now be
converted. This covers all spots found by "make coccicheck"
in master (the coccicheck results were tweaked by hand for
style).

Signed-off-by: Jeff King 
---
Jake: I was surprised that this was not a "patch 2" on top of your
earlier coccicheck patch. Apologies if you were planning to send it out.

This doesn't conflict with anything in "pu", so it's a reasonable time
to apply it. There are a few lingering cases in pu, so another option is
to wait a week or two and see if they get merged.

 builtin/checkout.c | 3 ++-
 cache-tree.c   | 2 +-
 commit-reach.c | 2 +-
 midx.c | 8 
 4 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/builtin/checkout.c b/builtin/checkout.c
index b30b48767e..902c06702c 100644
--- a/builtin/checkout.c
+++ b/builtin/checkout.c
@@ -497,7 +497,8 @@ static int skip_merge_working_tree(const struct 
checkout_opts *opts,
 * We must do the merge if we are actually moving to a new commit.
 */
if (!old_branch_info->commit || !new_branch_info->commit ||
-   oidcmp(_branch_info->commit->object.oid, 
_branch_info->commit->object.oid))
+   !oideq(_branch_info->commit->object.oid,
+  _branch_info->commit->object.oid))
return 0;
 
/*
diff --git a/cache-tree.c b/cache-tree.c
index 5ce51468f0..9c5cf2cc4f 100644
--- a/cache-tree.c
+++ b/cache-tree.c
@@ -781,7 +781,7 @@ static void verify_one(struct index_state *istate,
strbuf_add(_buf, oid->hash, the_hash_algo->rawsz);
}
hash_object_file(tree_buf.buf, tree_buf.len, tree_type, _oid);
-   if (oidcmp(_oid, >oid))
+   if (!oideq(_oid, >oid))
BUG("cache-tree for path %.*s does not match. "
"Expected %s got %s", len, path->buf,
oid_to_hex(_oid), oid_to_hex(>oid));
diff --git a/commit-reach.c b/commit-reach.c
index 00e5ceee6f..a7808430e1 100644
--- a/commit-reach.c
+++ b/commit-reach.c
@@ -426,7 +426,7 @@ struct contains_stack {
 static int in_commit_list(const struct commit_list *want, struct commit *c)
 {
for (; want; want = want->next)
-   if (!oidcmp(>item->object.oid, >object.oid))
+   if (oideq(>item->object.oid, >object.oid))
return 1;
return 0;
 }
diff --git a/midx.c b/midx.c
index f3e8dbc108..06aef56400 100644
--- a/midx.c
+++ b/midx.c
@@ -285,8 +285,8 @@ static int nth_midxed_pack_entry(struct multi_pack_index 
*m, struct pack_entry *
struct object_id oid;
nth_midxed_object_oid(, m, pos);
for (i = 0; i < p->num_bad_objects; i++)
-   if (!hashcmp(oid.hash,
-p->bad_object_sha1 + the_hash_algo->rawsz 
* i))
+   if (hasheq(oid.hash,
+  p->bad_object_sha1 + the_hash_algo->rawsz * 
i))
return 0;
}
 
@@ -583,8 +583,8 @@ static struct pack_midx_entry *get_sorted_entries(struct 
multi_pack_index *m,
 * Take only the first duplicate.
 */
for (cur_object = 0; cur_object < nr_fanout; cur_object++) {
-   if (cur_object && !oidcmp(_by_fanout[cur_object 
- 1].oid,
- 
_by_fanout[cur_object].oid))
+   if (cur_object && oideq(_by_fanout[cur_object - 
1].oid,
+   
_by_fanout[cur_object].oid))
continue;
 
ALLOC_GROW(deduplicated_entries, *nr_objects + 1, 
alloc_objects);
-- 
2.19.0.489.g7c40b6af8e