Hello, On 16.11.20 09:21, Sascha Hauer wrote: > On Fri, Nov 13, 2020 at 09:26:00AM +0100, Ahmad Fatoum wrote: >> Previously setenv xxx= deleted x like xxx= does. With xxx= now instead >> assigning the empty string, setenv xxx= should too. >> >> Signed-off-by: Ahmad Fatoum <a.fatoum@xxxxxxxxxxxxxx> >> >> --- >> commands/setenv.c | 3 +-- >> 1 file changed, 1 insertion(+), 2 deletions(-) >> >> diff --git a/commands/setenv.c b/commands/setenv.c >> index 6992f604f513..9aeb8f010bc5 100644 >> --- a/commands/setenv.c >> +++ b/commands/setenv.c >> @@ -18,8 +18,7 @@ static int do_setenv(int argc, char *argv[]) >> equal = strrchr(argv[1], '='); >> if (equal) { >> equal[0] = '\0'; >> - if (equal[1]) >> - argv[2] = &equal[1]; >> + argv[2] = &equal[1]; >> } > > Fine with me, but I just had a second look at fbf145dc86 ("commands: setenv: > support setenv dev.var=VAL syntax"). This patch introduced using argv[2] > when argc was only to, thus only argv[0] and argv[1] are valid. I wonder > if this is allowed. KAsan doesn't complain, but still I think this > should be changed. Could you do this before further modifying this? barebox always has argv[argc] = NULL, so just like standard C, accessing argv[2] here is fine. See common/parser.c:29: argv[nargs] = NULL; common/hush.c:698: pglob->gl_pathv[pathc] = NULL; lib/glob.c:398: pglob->gl_pathv[pglob->gl_pathc] = NULL; lib/glob.c:232: pglob->gl_pathv[pglob->gl_pathc] = NULL; Cheers, Ahmad > > Sascha > -- Pengutronix e.K. | | Steuerwalder Str. 21 | http://www.pengutronix.de/ | 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 | _______________________________________________ barebox mailing list barebox@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/barebox