On 10/11/16 20:03, peng.liang5@xxxxxxxxxx wrote: > From: peng liang <peng.liang5@xxxxxxxxxx> > > -add missing newline to 'map|multipath $map getprstatus' reply > -use asprintf instead of sprintf > > Signed-off-by: peng liang <peng.liang5@xxxxxxxxxx> > --- > multipathd/cli_handlers.c | 14 ++++++-------- > 1 file changed, 6 insertions(+), 8 deletions(-) > > diff --git a/multipathd/cli_handlers.c b/multipathd/cli_handlers.c > index 8ff4362..16445ea 100644 > --- a/multipathd/cli_handlers.c > +++ b/multipathd/cli_handlers.c > @@ -1,6 +1,9 @@ > /* > * Copyright (c) 2005 Christophe Varoqui > */ > +#define _GNU_SOURCE > + > +#include <stdio.h> > #include "checkers.h" > #include "memory.h" > #include "vector.h" > @@ -1285,14 +1288,9 @@ cli_getprstatus (void * v, char ** reply, int * > len, void * data) > > condlog(3, "%s: prflag = %u", param, (unsigned int)mpp->prflag); > > - *reply =(char *)malloc(2); > - *len = 2; > - memset(*reply,0,2); > - > - > - sprintf(*reply,"%d",mpp->prflag); > - (*reply)[1]='\0'; > - > + *len = asprintf(reply, "%d\n", mpp->prflag); > + if (*len < 0) > + return 1; > > condlog(3, "%s: reply = %s", param, *reply); Hello Peng, Sorry but returning 1 looks somewhat inconsistent to me. This function is called indirectly by uxsock_trigger() and that function expects that cli_getprstatus() either returns a positive error code (E...) or a negative error code. Please change this patch such that ENOMEM is returned instead of 1 if asprintf() fails. Thanks, Bart. -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
