From: Johannes Schindelin <Johannes.Schindelin@gmx.de>
To: Matheus Tavares <matheus.bernardino@usp.br>
Cc: git@vger.kernel.org, sandals@crustytoothpaste.net, j6t@kdbg.org,
jonathantanmy@google.com, peff@peff.net,
christian.couder@gmail.com, Fredrik Kuivinen <frekui@gmail.com>
Subject: Re: [PATCH v2 2/2] hex: make hash_to_hex_algop() and friends thread-safe
Date: Thu, 16 Jul 2020 13:29:52 +0200 (CEST) [thread overview]
Message-ID: <nycvar.QRO.7.76.6.2007161214270.54@tvgsbejvaqbjf.bet> (raw)
In-Reply-To: <b47445fa1cef6d4523dd0ca336f7ee22bce89466.1593208411.git.matheus.bernardino@usp.br>
Hi Matheus,
On Fri, 26 Jun 2020, Matheus Tavares wrote:
> @@ -136,12 +137,49 @@ char *oid_to_hex_r(char *buffer, const struct object_id *oid)
> return hash_to_hex_algop_r(buffer, oid->hash, the_hash_algo);
> }
>
> +struct hexbuf_array {
> + int idx;
> + char bufs[4][GIT_MAX_HEXSZ + 1];
> +};
> +
> +#ifdef HAVE_THREADS
> +static pthread_key_t hexbuf_array_key;
> +static pthread_once_t hexbuf_array_once = PTHREAD_ONCE_INIT;
> +
> +static void init_hexbuf_array_key(void)
> +{
> + if (pthread_key_create(&hexbuf_array_key, free))
> + die(_("failed to initialize threads' key for hash to hex conversion"));
> +}
> +
> +#else
> +static struct hexbuf_array default_hexbuf_array;
> +#endif
> +
> char *hash_to_hex_algop(const unsigned char *hash, const struct git_hash_algo *algop)
> {
> - static int bufno;
> - static char hexbuffer[4][GIT_MAX_HEXSZ + 1];
> - bufno = (bufno + 1) % ARRAY_SIZE(hexbuffer);
> - return hash_to_hex_algop_r(hexbuffer[bufno], hash, algop);
> + struct hexbuf_array *ha;
> +
> +#ifdef HAVE_THREADS
> + void *value;
> +
> + if (pthread_once(&hexbuf_array_once, init_hexbuf_array_key))
> + die(_("failed to initialize threads' key for hash to hex conversion"));
> +
> + value = pthread_getspecific(hexbuf_array_key);
> + if (value) {
> + ha = (struct hexbuf_array *) value;
> + } else {
> + ha = xmalloc(sizeof(*ha));
I just realized (while trying to debug something independent) that this
leaves `ha->idx` uninitialized. So you will need at least this patch to
fix a bug that currently haunts `seen`'s CI builds (you can use
`--valgrind`, like I did, to identify the problem):
-- snip --
diff --git a/hex.c b/hex.c
index 4f2f163d5e7..365ba94ab11 100644
--- a/hex.c
+++ b/hex.c
@@ -171,6 +171,7 @@ char *hash_to_hex_algop(const unsigned char *hash, const struct git_hash_algo *a
ha = (struct hexbuf_array *) value;
} else {
ha = xmalloc(sizeof(*ha));
+ ha->idx = 0;
if (pthread_setspecific(hexbuf_array_key, (void *)ha))
die(_("failed to set thread buffer for hash to hex conversion"));
}
-- snap --
But as I mentioned before, I would be much more in favor of abandoning
this thread-local idea (because it is _still_ fragile, as the same thread
could try to make use of more than four hex values in the same `printf()`,
for example) and instead using Coccinelle to convert all those
`oid_to_hex()` calls to `oid_to_hex_r()` calls.
Now, I am _far_ from knowing what I'm doing with Coccinelle, but I think
this here semantic patch should get you going:
-- snipsnap --
@@
expression E;
@@
{
++ char hex[GIT_MAX_HEXSZ + 1];
...
- oid_to_hex(E)
+ oid_to_hex_r(hex, E)
...
}
@@
expression E1, E2;
@@
{
++ char hex1[GIT_MAX_HEXSZ + 1], hex2[GIT_MAX_HEXSZ + 1];
...
- oid_to_hex(E1)
+ oid_to_hex_r(hex1, E1)
...
- oid_to_hex(E2)
+ oid_to_hex_r(hex2, E2)
...
}
next prev parent reply other threads:[~2020-07-16 12:56 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-24 22:29 [RFC] Thread safety in some low-level functions Matheus Tavares Bernardino
2020-06-24 22:52 ` Matheus Tavares Bernardino
2020-06-25 1:38 ` brian m. carlson
2020-06-25 20:32 ` [PATCH 0/2] Make oid_to_hex() thread-safe Matheus Tavares
2020-06-25 20:32 ` [PATCH 1/2] compat/win32/pthread: add pthread_once() Matheus Tavares
2020-06-26 5:45 ` Christian Couder
2020-06-26 14:13 ` Matheus Tavares Bernardino
2020-06-25 20:32 ` [PATCH 2/2] hex: make hash_to_hex_algop() and friends thread-safe Matheus Tavares
2020-06-25 23:07 ` brian m. carlson
2020-06-25 23:54 ` Matheus Tavares Bernardino
2020-06-26 0:00 ` Matheus Tavares Bernardino
2020-06-26 6:02 ` Christian Couder
2020-06-26 8:22 ` [PATCH 0/2] Make oid_to_hex() thread-safe Christian Couder
2020-06-26 16:22 ` Matheus Tavares Bernardino
2020-06-26 21:54 ` [PATCH v2 " Matheus Tavares
2020-06-26 21:54 ` [PATCH v2 1/2] compat/win32/pthread: add pthread_once() Matheus Tavares
2020-06-26 21:54 ` [PATCH v2 2/2] hex: make hash_to_hex_algop() and friends thread-safe Matheus Tavares
2020-06-29 15:11 ` Johannes Schindelin
2020-06-30 20:37 ` Matheus Tavares Bernardino
2020-07-01 11:32 ` Johannes Schindelin
2020-07-16 11:29 ` Johannes Schindelin [this message]
2020-07-18 3:09 ` Matheus Tavares Bernardino
2020-08-10 14:15 ` Johannes Schindelin
2020-07-18 3:52 ` Matheus Tavares
2020-07-26 17:41 ` René Scharfe
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=nycvar.QRO.7.76.6.2007161214270.54@tvgsbejvaqbjf.bet \
--to=johannes.schindelin@gmx.de \
--cc=christian.couder@gmail.com \
--cc=frekui@gmail.com \
--cc=git@vger.kernel.org \
--cc=j6t@kdbg.org \
--cc=jonathantanmy@google.com \
--cc=matheus.bernardino@usp.br \
--cc=peff@peff.net \
--cc=sandals@crustytoothpaste.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).