linux-sparse.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Phil Carmody <phil@dovecot.fi>
To: sparse@chrisli.org
Cc: josh@joshtriplett.org, linux-sparse@vger.kernel.org, phil@dovecot.fi
Subject: [PATCHv2 2/3] sparse: detect non-sign-extended masks created by '~'
Date: Tue, 10 Jun 2014 10:54:06 +0300	[thread overview]
Message-ID: <1402386847-23477-3-git-send-email-phil@dovecot.fi> (raw)
In-Reply-To: <1402386847-23477-2-git-send-email-phil@dovecot.fi>

Consider the operation of rounding up to the nearest multiple of a power of 2.
e.g.  #define ALLOC_SIZE(t) ((sizeof(t) + ASIZE - 1) & ~(ASIZE - 1))

If ASIZE is unfortunately defined as an unsigned type smaller than size_t,
then the ~ will not undergo sign-bit extension, and an incorrect mask will
be used. If used in a memory allocation context this could be fatal.

Warn about such dubious 'large op ~short' usage.

v2: pulled noisy repeated parts into a helper

Signed-off-by: Phil Carmody <phil@dovecot.fi>
---
 evaluate.c | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/evaluate.c b/evaluate.c
index 9052962..a16aa45 100644
--- a/evaluate.c
+++ b/evaluate.c
@@ -886,6 +886,25 @@ static struct symbol *evaluate_logical(struct expression *expr)
 	return &int_ctype;
 }
 
+static int int_size_cmp(struct symbol *left, struct symbol *right)
+{
+	left = integer_promotion(left);
+	right = integer_promotion(right);
+
+	return (left->bit_size > right->bit_size) ? 1 :
+		(right->bit_size > left->bit_size) ? -1 : 0;
+}
+
+static void check_masking(struct expression *expr, int op, int is_l,
+			  struct expression *sexpr, struct symbol *stype)
+{
+	if ((sexpr->type == EXPR_PREOP)
+	    && (sexpr->op == '~')
+	    && (stype->ctype.modifiers & MOD_UNSIGNED))
+		warning(expr->pos, "dubious zero-extended '~': %sx %c %sy",
+			"~"+!is_l, op, "~"+!!is_l);
+}
+
 static struct symbol *evaluate_binop(struct expression *expr)
 {
 	struct symbol *ltype, *rtype, *ctype;
@@ -917,6 +936,7 @@ static struct symbol *evaluate_binop(struct expression *expr)
 			rtype = integer_promotion(rtype);
 		} else {
 			// The rest do usual conversions
+			int size_cmp;
 			int left_not  = expr->left->type == EXPR_PREOP
 					&& expr->left->op == '!';
 			int right_not = expr->right->type == EXPR_PREOP
@@ -927,6 +947,12 @@ static struct symbol *evaluate_binop(struct expression *expr)
 					op,
 					right_not ? "!" : "");
 
+			size_cmp = int_size_cmp(ltype, rtype);
+			if (size_cmp > 0)
+				check_masking(expr, op, 0, expr->right, rtype);
+			else if (size_cmp < 0)
+				check_masking(expr, op, 1, expr->left, ltype);
+
 			ltype = usual_conversions(op, expr->left, expr->right,
 						  lclass, rclass, ltype, rtype);
 			ctype = rtype = ltype;
-- 
2.0.0


  reply	other threads:[~2014-06-10  7:53 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-06-09 11:57 [PATCH 0/3] catch non-sign-extended '~' brainos Phil Carmody
2014-06-09 11:58 ` [PATCH 1/3] sparse: Just use simple ints for decision variables Phil Carmody
2014-06-09 11:58   ` [PATCH 2/3] sparse: detect non-sign-extended masks created by '~' Phil Carmody
2014-06-09 11:58     ` [PATCH 3/3] validation: dubious bitwise operations with nots Phil Carmody
2014-06-09 13:36       ` Josh Triplett
2014-06-09 13:34     ` [PATCH 2/3] sparse: detect non-sign-extended masks created by '~' Josh Triplett
2014-06-09 16:05       ` Phil Carmody
2014-06-09 13:27   ` [PATCH 1/3] sparse: Just use simple ints for decision variables Josh Triplett
2014-06-10  7:54 ` [PATCHv2 0/3] catch non-sign-extended '~' brainos Phil Carmody
2014-06-10  7:54   ` [PATCHv2 1/3] sparse: Just use simple ints for decision variables Phil Carmody
2014-06-10  7:54     ` Phil Carmody [this message]
2014-06-10  7:54       ` [PATCHv2 3/3] validation: dubious bitwise operations with bitwise nots Phil Carmody
2014-06-27 11:19   ` [PATCHv2 0/3] catch non-sign-extended '~' brainos Phil Carmody
2014-06-27 17:16     ` Christopher Li
2014-06-30  8:56       ` Phil Carmody
     [not found]         ` <CANeU7Q=Z=Xac_T3JRAyqo_fF4LAKD-MM41NYz+nDstDutcVUfA@mail.gmail.com>
2014-06-30 17:27           ` Christopher Li
2014-07-01 11:30           ` Phil Carmody
2014-07-01 19:42             ` Christopher Li
2014-07-02  7:43               ` Phil Carmody
2014-07-02  8:51                 ` Christopher Li
2014-07-02  9:28                   ` Phil Carmody

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=1402386847-23477-3-git-send-email-phil@dovecot.fi \
    --to=phil@dovecot.fi \
    --cc=josh@joshtriplett.org \
    --cc=linux-sparse@vger.kernel.org \
    --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 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 NNTP newsgroup(s).