git/list[1] front-page[2] threads[3] people[4] search[5] about
 

[PATCH 1/2] config: properly range-check integer values

From
Jeff King <peff@peff.net>
Date
Aug 20, 2013, 22:42 UTC
Message-ID
<20130820224256.GA24766@sigill.intra.peff.net>
In-Reply-To
<20130820223953.GA3429@sigill.intra.peff.net>

When we look at a config value as an integer using the git_config_int function, we carefully range-check the value we get and complain if it is out of our range. But the range we compare to is that of a "long", which we then cast to an "int" in the function's return value. This means that on systems where "int" and "long" have different sizes (e.g., LP64 systems), we may pass the range check, but then return nonsense by truncating the value as we cast it to an int.

We can solve this by converting git_parse_long into git_parse_int, and range-checking the "int" range. Nobody actually cared that we used a "long" internally, since the result was truncated anyway, and there were no other callers of git_parse_long.

Of the existing callers, all of them expect small-ish values that are fine for an "int", and should not be impacted except when nonsense is fed to them. The one arguable exception is http.lowSpeedLimit, which is given in bytes per second, and is fed to curl as a long. However, nobody is likely to consider 2GB/sec as "low speed", and even if they did, the new behavior is much better (it barfs and complains rather than considering the value negative and silently ignoring it).

Signed-off-by: Jeff King <peff@peff.net>
---
I added a test. It would not fail on existing 32-bit systems, but would
on existing LP64 systems. It will pass with the new code on both.
However, it will fail on ILP64 systems (because their int is large, and
can represent 3GB). I'm not sure if any such systems are in wide use
(SPARC64?), but we would want a prereq in that case, I guess. I'm
inclined to wait to see if it actually fails for anybody.
 config.c               | 12 ++++++------
 t/t1300-repo-config.sh |  5 +++++
 2 files changed, 11 insertions(+), 6 deletions(-)
diff --git a/config.c b/config.c
index e13a7b6..9237d4b 100644
--- a/config.c
+++ b/config.c
@@ -468,7 +468,7 @@ static int parse_unit_factor(const char *end, uintmax_t *val)
 	return 0;
 }
 
-static int git_parse_long(const char *value, long *ret)
+static int git_parse_int(const char *value, int *ret)
 {
 	if (value && *value) {
 		char *end;
@@ -484,7 +484,7 @@ static int git_parse_long(const char *value, long *ret)
 			return 0;
 		uval = abs(val);
 		uval *= factor;
-		if ((uval > maximum_signed_value_of_type(long)) ||
+		if ((uval > maximum_signed_value_of_type(int)) ||
 		    (abs(val) > uval))
 			return 0;
 		val *= factor;
@@ -526,8 +526,8 @@ int git_config_int(const char *name, const char *value)
 
 int git_config_int(const char *name, const char *value)
 {
-	long ret = 0;
-	if (!git_parse_long(value, &ret))
+	int ret;
+	if (!git_parse_int(value, &ret))
 		die_bad_config(name);
 	return ret;
 }
@@ -559,10 +559,10 @@ int git_config_maybe_bool(const char *name, const char *value)
 
 int git_config_maybe_bool(const char *name, const char *value)
 {
-	long v = git_config_maybe_bool_text(name, value);
+	int v = git_config_maybe_bool_text(name, value);
 	if (0 <= v)
 		return v;
-	if (git_parse_long(value, &v))
+	if (git_parse_int(value, &v))
 		return !!v;
 	return -1;
 }
diff --git a/t/t1300-repo-config.sh b/t/t1300-repo-config.sh
index c4a7d84..3c5ec4b 100755
--- a/t/t1300-repo-config.sh
+++ b/t/t1300-repo-config.sh
@@ -652,6 +652,11 @@ test_expect_success numbers '
 	test_cmp expect actual
 '
 
+test_expect_success 'out-of-range integers are detected' '
+	git config giga.crash 3g &&
+	test_must_fail git config --int giga.crash
+'
+
 test_expect_success 'invalid unit' '
 	git config aninvalid.unit "1auto" &&
 	echo 1auto >expect &&
-- 
1.8.4.rc2.28.g6bb5f3f
Previous: Jeff KingNext: Jonathan Nieder
Message 2 of 25 in “git-config and large integers”
  1. 0/2 git-config and large integersJeff King, Aug 20, 2013
  2. 1/2 config: properly range-check integer valuesJeff King, Aug 20, 2013
  3. Jonathan NiederAug 20, 2013
  4. Jeff KingAug 21, 2013
  5. Stefan BellerAug 20, 2013
  6. Jeff KingAug 20, 2013
  7. 2/2 teach git-config to output large integersJeff King, Aug 20, 2013
  8. Jonathan NiederAug 20, 2013
  9. Jeff KingAug 21, 2013
  10. Jonathan NiederAug 21, 2013
  11. Jeff KingAug 21, 2013
  12. Jonathan NiederAug 21, 2013
  13. Junio C HamanoAug 20, 2013
  14. Junio C HamanoAug 20, 2013
  15. Jeff KingAug 21, 2013
  16. Jeff KingAug 21, 2013
  17. 0/5 git-config and large integersJeff King, Sep 8, 2013
  18. 1/5 config: factor out integer parsing from range checksJeff King, Sep 8, 2013
  19. 2/5 config: properly range-check integer valuesJeff King, Sep 8, 2013
  20. 3/5 config: set errno in numeric git_parse_* functionsJeff King, Sep 8, 2013
  21. Eric SunshineSep 9, 2013
  22. Jeff KingSep 9, 2013
  23. 4/5 config: make numeric parsing errors more clearJeff King, Sep 8, 2013
  24. 5/5 git-config: always treat --int as 64-bit internallyJeff King, Sep 8, 2013
  25. Junio C HamanoSep 9, 2013

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.