Greetings, The dmsetup tool currently has a limit of 1024 characters for the length of a line in a table file when running the 'create' or 'load' command. When have run into scenarios where we've exceeded this length, usually due to creating large striped devices with unusually long names for the underlying disks. To get around this limitation, I wrote a patch that simply changes from using fgets() to getline() in the _parse_file() routine. The getline() routine will automatically allocate (or reallocate) space as needed to read in an entire line from a file. I have tested this patch successfully with table lines of at least 1300 characters. Thanks, -- Kevin Corry kevcorry@xxxxxxxxxx http://www.ibm.com/linux/ http://evms.sourceforge.net/ Use getline() instead of fgets() to remove the line-size limitation when parsing the input file in dmsetup. Signed-Off-By: Kevin Corry <kevcorry@xxxxxxxxxx> Index: device-mapper.1.02.05/dmsetup/dmsetup.c =================================================================== --- device-mapper.1.02.05.orig/dmsetup/dmsetup.c +++ device-mapper.1.02.05/dmsetup/dmsetup.c @@ -117,7 +117,8 @@ static struct dm_tree *_dtree; */ static int _parse_file(struct dm_task *dmt, const char *file) { - char buffer[LINE_SIZE], ttype[LINE_SIZE], *ptr, *comment; + char *buffer = NULL, ttype[LINE_SIZE], *ptr, *comment; + size_t buffer_size = 0; FILE *fp; unsigned long long start, size; int r = 0, n, line = 0; @@ -132,7 +133,7 @@ static int _parse_file(struct dm_task *d } else fp = stdin; - while (fgets(buffer, sizeof(buffer), fp)) { + while (getline(&buffer, &buffer_size, fp) > 0) { line++; /* trim trailing space */ @@ -166,6 +167,7 @@ static int _parse_file(struct dm_task *d out: if (file) fclose(fp); + free(buffer); return r; } -- dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
