linux-sparse.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
To: linux-sparse@vger.kernel.org
Cc: Christopher Li <sparse@chrisli.org>,
	Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Subject: [PATCH 4/5] do not report bad types twice or more
Date: Tue, 19 Sep 2017 04:13:34 +0200	[thread overview]
Message-ID: <20170919021335.5881-5-luc.vanoostenryck@gmail.com> (raw)
In-Reply-To: <20170919021335.5881-1-luc.vanoostenryck@gmail.com>

The type 'bad_ctype' is only used after an error has been detected.
Since this error has also been reported, there is no reasons
to issue more errors when a 'bad_ctype' is involved. This allow
to focus on the root cause of the error.

Fix this by checking in bad_expr_type() if one of the operands
is already a 'bad_ctype' and do not issue an diagnostic message
in this case.

Note: the kernel has a bunch of these situation where sometimes
      the exact same warning is given several times in a row,
      sometimes as much as a dozen time.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
---
 evaluate.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/evaluate.c b/evaluate.c
index 0394dcbb9..c16ee9624 100644
--- a/evaluate.c
+++ b/evaluate.c
@@ -400,15 +400,20 @@ static inline int is_string_type(struct symbol *type)
 
 static struct symbol *bad_expr_type(struct expression *expr)
 {
-	sparse_error(expr->pos, "incompatible types for operation (%s)", show_special(expr->op));
 	switch (expr->type) {
 	case EXPR_BINOP:
 	case EXPR_COMPARE:
+		if (!valid_expr_subtype(expr))
+			break;
+		sparse_error(expr->pos, "incompatible types for operation (%s)", show_special(expr->op));
 		info(expr->pos, "   left side has type %s", show_typename(expr->left->ctype));
 		info(expr->pos, "   right side has type %s", show_typename(expr->right->ctype));
 		break;
 	case EXPR_PREOP:
 	case EXPR_POSTOP:
+		if (!valid_type(expr->unop->ctype))
+			break;
+		sparse_error(expr->pos, "incompatible types for operation (%s)", show_special(expr->op));
 		info(expr->pos, "   argument has type %s", show_typename(expr->unop->ctype));
 		break;
 	default:
@@ -887,6 +892,8 @@ static struct symbol *evaluate_conditional(struct expression *expr, int iterator
 			if (Waddress)
 				warning(expr->pos, "the address of %s will always evaluate as true", "an array");
 		} else if (!is_scalar_type(ctype)) {
+			if (!valid_type(ctype))
+				return NULL;
 			sparse_error(expr->pos, "incorrect type in conditional");
 			info(expr->pos, "   got %s", show_typename(ctype));
 			ctype = NULL;
-- 
2.14.0


  parent reply	other threads:[~2017-09-19  2:13 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-09-19  2:13 [PATCH 0/5] expressions without a type Luc Van Oostenryck
2017-09-19  2:13 ` [PATCH 1/5] do not linearize " Luc Van Oostenryck
2017-09-19  2:13 ` [PATCH 2/5] add helper: valid_type() Luc Van Oostenryck
2017-09-19  2:13 ` [PATCH 3/5] add helper: valid_expr_subtype() Luc Van Oostenryck
2017-09-19  2:13 ` Luc Van Oostenryck [this message]
2017-09-19  2:13 ` [PATCH 5/5] always evaluate both operands 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=20170919021335.5881-5-luc.vanoostenryck@gmail.com \
    --to=luc.vanoostenryck@gmail.com \
    --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).