All of lore.kernel.org
 help / color / mirror / Atom feed
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
To: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: linux-sparse@vger.kernel.org, Christopher Li <sparse@chrisli.org>
Subject: [PATCH] Avoid reuse of string buffer when concatening adjacent string litterals
Date: Sat, 31 Jan 2015 02:23:40 +0100	[thread overview]
Message-ID: <20150131012339.GA3460@macpro.local> (raw)
In-Reply-To: <87y4ojhq2f.fsf@rasmusvillemoes.dk>

In get_string_constant(), the code tried to reuse the storage for the string
but only if the expansion of the string was not bigger than its unexpanded form.
But this fail when the string constant is a sequence of adjacent string litterals
(each being possibly shared, used elsewhere, isolated or in another order).
The minimal exemple would be something like this:

#define P "\001"
const char a[] = P "a";
const char b[] = P "b";

The expansion for 'a' will produce a string which is smaller than
the unexpanded "\001" (2 instead of 4).
By trying to reuse the storage, all further occurrence of "\001"
(probably only from the same 'origin', here the macro P) will then be replaced by "\001a".

The fix is thus to not try to reuse the storage for the string if it consit of
several adjacent litterals.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
---
 char.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/char.c b/char.c
index 08ca2230..ce1a0700 100644
--- a/char.c
+++ b/char.c
@@ -93,6 +93,7 @@ struct token *get_string_constant(struct token *token, struct expression *expr)
 	static char buffer[MAX_STRING];
 	int len = 0;
 	int bits;
+	int parts = 0;
 
 	while (!done) {
 		switch (token_type(next)) {
@@ -117,13 +118,14 @@ struct token *get_string_constant(struct token *token, struct expression *expr)
 			len++;
 		}
 		token = token->next;
+		parts++;
 	}
 	if (len > MAX_STRING) {
 		warning(token->pos, "trying to concatenate %d-character string (%d bytes max)", len, MAX_STRING);
 		len = MAX_STRING;
 	}
 
-	if (len >= string->length)	/* can't cannibalize */
+	if (len >= string->length || parts > 1)	/* safe to reuse the string buffer */
 		string = __alloc_string(len+1);
 	string->length = len+1;
 	memcpy(string->data, buffer, len);
-- 
2.2.2



  reply	other threads:[~2015-01-31  1:23 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-01-30 22:16 Bad interaction between macro expansion and literal concatenation Rasmus Villemoes
2015-01-31  1:23 ` Luc Van Oostenryck [this message]
2015-02-03 22:38   ` [PATCH] Avoid reuse of string buffer when concatening adjacent string litterals Rasmus Villemoes
2015-02-04  0:32     ` Luc Van Oostenryck
2015-02-04  3:26       ` Christopher Li
2015-02-04  8:39       ` Rasmus Villemoes
2015-02-04  8:58         ` Rasmus Villemoes
2015-02-04 16:20           ` Christopher Li
2015-02-06 21:52             ` Rasmus Villemoes
2015-02-07  1:30               ` Christopher Li
2015-02-09 21:48                 ` Damien Lespiau
2015-02-04  2:01     ` [PATCH v2] Avoid reusing string buffer when doing string expansion Luc Van Oostenryck
2015-02-04  5:30       ` Christopher Li
2015-02-04  6:22         ` Luc Van Oostenryck
2015-02-04  8:01           ` Christopher Li
2015-02-04 16:38             ` Christopher Li
2015-02-04 23:38               ` Luc Van Oostenryck
2015-02-06 13:58                 ` Christopher Li
2015-02-06 20:32                   ` Rasmus Villemoes
2015-02-04 23:38             ` Luc Van Oostenryck
2015-01-31  5:16 ` Bad interaction between macro expansion and literal concatenation Christopher Li
2015-02-01  2:19   ` [PATCH 0/4] Teach sparse to display data/initial values Luc Van Oostenryck
2015-02-01  2:19     ` [PATCH 1/4] Add support for '-vdata', the equivalent of '-ventry' but for data Luc Van Oostenryck
2015-02-01  2:19     ` [PATCH 2/4] Add support for show_data() Luc Van Oostenryck
2015-02-02  5:30       ` Christopher Li
2015-02-04  0:50         ` Luc Van Oostenryck
2015-02-01  2:19     ` [PATCH 3/4] Teach sparse to display data/initial values Luc Van Oostenryck
2015-02-01  2:19     ` [PATCH 4/4] Small test/exemple for using '-vdata' Luc Van Oostenryck

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

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20150131012339.GA3460@macpro.local \
    --to=luc.vanoostenryck@gmail.com \
    --cc=linux-sparse@vger.kernel.org \
    --cc=linux@rasmusvillemoes.dk \
    --cc=sparse@chrisli.org \
    /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.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.