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

[PATCH] diff: use 64-bit integers for diffstat calculations

From
Jeff King <peff@peff.net>
Date
Apr 17, 2010, 10:25 UTC
Message-ID
<20100417102543.GB23110@coredump.intra.peff.net>
In-Reply-To
<4BC87BE9.9040704@dbservice.com>
On Fri, Apr 16, 2010 at 05:02:01PM +0200, Tomas Carnecky wrote:
Show 12 quoted lines
> >commit 6afe3d3c889daa92bd79956c4bb733eb5cb408dc
> >Author: Heikki Orsila<heikki.orsila@iki.fi>
> >Date:   2010-04-16 16:54:52 +0300
> >
> >     test commit
> >
> >  0 |  Bin 0 ->  -2146435072 bytes
> >  1 |  Bin 0 ->  -2146435072 bytes
> >  2 files changed, 0 insertions(+), 0 deletions(-)
> 
> Yep, bug is also in the latest version (1.7.0.5). The code uses 'int'
> instead of something big enough to hold the size of your files.

Yuck, we use "unsigned int" for the actual storage, and then convert to a regular "int" in some other places. I think we should just do this:

-- >8 --
Subject: [PATCH] diff: use 64-bit integers for diffstat calculations

The diffstat "added" and "changed" fields generally store line counts; however, for binary files, they store file sizes. Since we store and print these values as ints, a diffstat on a file larger than 2G can show a negative size. Instead, let's explicitly use 64-bit integers.

Signed-off-by: Jeff King <peff@peff.net>
---
 diff.c |   21 ++++++++++++---------
 1 files changed, 12 insertions(+), 9 deletions(-)
diff --git a/diff.c b/diff.c
index 7effdac..54933cc 100644
--- a/diff.c
+++ b/diff.c
@@ -936,7 +936,7 @@ struct diffstat_t {
 		unsigned is_unmerged:1;
 		unsigned is_binary:1;
 		unsigned is_renamed:1;
-		unsigned int added, deleted;
+		uint64_t added, deleted;
 	} **files;
 };
 
@@ -1028,7 +1028,7 @@ static void fill_print_name(struct diffstat_file *file)
 static void show_stats(struct diffstat_t *data, struct diff_options *options)
 {
 	int i, len, add, del, adds = 0, dels = 0;
-	int max_change = 0, max_len = 0;
+	uint64_t max_change = 0, max_len = 0;
 	int total_files = data->nr;
 	int width, name_width;
 	const char *reset, *set, *add_c, *del_c;
@@ -1057,7 +1057,7 @@ static void show_stats(struct diffstat_t *data, struct diff_options *options)
 
 	for (i = 0; i < data->nr; i++) {
 		struct diffstat_file *file = data->files[i];
-		int change = file->added + file->deleted;
+		uint64_t change = file->added + file->deleted;
 		fill_print_name(file);
 		len = strlen(file->print_name);
 		if (max_len < len)
@@ -1085,8 +1085,8 @@ static void show_stats(struct diffstat_t *data, struct diff_options *options)
 	for (i = 0; i < data->nr; i++) {
 		const char *prefix = "";
 		char *name = data->files[i]->print_name;
-		int added = data->files[i]->added;
-		int deleted = data->files[i]->deleted;
+		uint64_t added = data->files[i]->added;
+		uint64_t deleted = data->files[i]->deleted;
 		int name_len;
 
 		/*
@@ -1107,9 +1107,11 @@ static void show_stats(struct diffstat_t *data, struct diff_options *options)
 		if (data->files[i]->is_binary) {
 			show_name(options->file, prefix, name, len);
 			fprintf(options->file, "  Bin ");
-			fprintf(options->file, "%s%d%s", del_c, deleted, reset);
+			fprintf(options->file, "%s%"PRIu64"%s",
+				del_c, deleted, reset);
 			fprintf(options->file, " -> ");
-			fprintf(options->file, "%s%d%s", add_c, added, reset);
+			fprintf(options->file, "%s%"PRIu64"%s",
+				add_c, added, reset);
 			fprintf(options->file, " bytes");
 			fprintf(options->file, "\n");
 			continue;
@@ -1138,7 +1140,7 @@ static void show_stats(struct diffstat_t *data, struct diff_options *options)
 			del = scale_linear(del, width, max_change);
 		}
 		show_name(options->file, prefix, name, len);
-		fprintf(options->file, "%5d%s", added + deleted,
+		fprintf(options->file, "%5"PRIu64"%s", added + deleted,
 				added + deleted ? " " : "");
 		show_graph(options->file, '+', add, add_c, reset);
 		show_graph(options->file, '-', del, del_c, reset);
@@ -1188,7 +1190,8 @@ static void show_numstat(struct diffstat_t *data, struct diff_options *options)
 			fprintf(options->file, "-\t-\t");
 		else
 			fprintf(options->file,
-				"%d\t%d\t", file->added, file->deleted);
+				"%"PRIu64"\t%"PRIu64"\t",
+				file->added, file->deleted);
 		if (options->line_termination) {
 			fill_print_name(file);
 			if (!file->is_renamed)
-- 
1.7.1.rc1.277.g2c4c9
Previous: Tomas CarneckyNext: Junio C Hamano
Message 3 of 5 in “A bug in git 1.6.5.2 with git log --stat: shows a negative number as a size”
  1. Heikki OrsilaApr 16, 2010
  2. Tomas CarneckyApr 16, 2010
  3. diff: use 64-bit integers for diffstat calculationsJeff King, Apr 17, 2010
  4. Junio C HamanoApr 17, 2010
  5. Jeff KingApr 17, 2010

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.