linux-kernel - [PATCH 4/7] perf trace: Rewind pointer in case field in header

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Thu, 13 May 2010 16:03:49 +1000
From:	"Ian Munsie" <imunsie@....ibm.com>
To:	linux-kernel@...r.kernel.org
Cc:	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Paul Mackerras <paulus@...ba.org>, Ingo Molnar <mingo@...e.hu>,
	Arnaldo Carvalho de Melo <acme@...hat.com>,
	Frederic Weisbecker <fweisbec@...il.com>,
	Steven Rostedt <rostedt@...dmis.org>,
	Ian Munsie <imunsie@....ibm.com>,
	Tom Zanussi <tzanussi@...il.com>
Subject: [PATCH 4/7] perf trace: Rewind pointer in case field in header_page is missing

From: Ian Munsie <imunsie@....ibm.com>

This patch rewinds the input pointer in the event that parsing one of
the fields in the header_page fails (which can happen if a field is
missing, such as the overwrite field), allowing the next field to be
parsed.

This patch fixes an issue from e9e94e3bd862d31777335722e747e97d9821bc1d
"perf trace: Ignore "overwrite" field if present in /events/header_page"
which ignored the field mismatch warning, but left the pointer in at
invalid position, resulting in this warning when the next field was
processed:

    	Warning: Error: expected type 6 but read 4

Signed-off-by: Ian Munsie <imunsie@....ibm.com>
---
 tools/perf/util/trace-event-parse.c |   35 +++++++++++++++++++++++------------
 1 files changed, 23 insertions(+), 12 deletions(-)

diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c
index 9717f7a..950c757 100644
--- a/tools/perf/util/trace-event-parse.c
+++ b/tools/perf/util/trace-event-parse.c
@@ -29,6 +29,7 @@
 #include <errno.h>
 
 #undef _GNU_SOURCE
+#include <linux/kernel.h>
 #include "../perf.h"
 #include "util.h"
 #include "trace-event.h"
@@ -37,6 +38,8 @@ int header_page_ts_offset;
 int header_page_ts_size;
 int header_page_size_offset;
 int header_page_size_size;
+/* NOTE: the overwrite field may not be present on all systems so, check if
+ * header_page_overwrite_size > 0 before use */
 int header_page_overwrite_offset;
 int header_page_overwrite_size;
 int header_page_data_offset;
@@ -3117,11 +3120,12 @@ static void parse_header_field(const char *field,
 {
 	char *token;
 	int type;
+	unsigned long long input_buf_ptr_sav = input_buf_ptr;
 
 	if (read_expected(EVENT_ITEM, "field") < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expected(EVENT_OP, ":") < 0)
-		return;
+		goto tokenlessfail;
 
 	/* type */
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
@@ -3129,29 +3133,29 @@ static void parse_header_field(const char *field,
 	free_token(token);
 
 	if (read_expected_warn(EVENT_ITEM, field, warn) < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expected(EVENT_OP, ";") < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expected(EVENT_ITEM, "offset") < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expected(EVENT_OP, ":") < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
 		goto fail;
 	*offset = atoi(token);
 	free_token(token);
 	if (read_expected(EVENT_OP, ";") < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expected(EVENT_ITEM, "size") < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expected(EVENT_OP, ":") < 0)
-		return;
+		goto tokenlessfail;
 	if (read_expect_type(EVENT_ITEM, &token) < 0)
 		goto fail;
 	*size = atoi(token);
 	free_token(token);
 	if (read_expected(EVENT_OP, ";") < 0)
-		return;
+		goto tokenlessfail;
 	type = read_token(&token);
 	if (type != EVENT_NEWLINE) {
 		/* newer versions of the kernel have a "signed" type */
@@ -3164,20 +3168,27 @@ static void parse_header_field(const char *field,
 		free_token(token);
 
 		if (read_expected(EVENT_OP, ":") < 0)
-			return;
+			goto tokenlessfail;
 
 		if (read_expect_type(EVENT_ITEM, &token))
 			goto fail;
 
 		free_token(token);
 		if (read_expected(EVENT_OP, ";") < 0)
-			return;
+			goto tokenlessfail;
 
 		if (read_expect_type(EVENT_NEWLINE, &token))
 			goto fail;
 	}
+	free_token(token);
+	return;
+
  fail:
 	free_token(token);
+ tokenlessfail:
+	input_buf_ptr = input_buf_ptr_sav;
+	if (warn)
+		pr_warning("Failed to parse %s field of header_page\n", field);
 }
 
 int parse_header_page(char *buf, unsigned long size)
-- 
1.7.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/