[PATCH] dump_log_entry: replace magic constant with a variable

[PATCH] dump_log_entry: replace magic constant with a variable

[Petr Tesarik]

If the buffer space requirements are changed, the magic constant is
hard to spot, and the failure is easily overlooked, because it may
fail only under special conditions.

Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
---
 makedumpfile.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/makedumpfile.c b/makedumpfile.c
index b4b6eca..37dea37 100644
--- a/makedumpfile.c
+++ b/makedumpfile.c
@@ -3949,7 +3949,7 @@ static int
 dump_log_entry(char *logptr, int fp)
 {
 	char *msg, *p, *bufp;
-	unsigned int i, text_len;
+	unsigned int i, text_len, buf_need;
 	unsigned long long ts_nsec;
 	char buf[BUFSIZE];
 	ulonglong nanos;
@@ -3966,9 +3966,11 @@ dump_log_entry(char *logptr, int fp)
 	bufp = buf;
 	bufp += sprintf(buf, "[%5lld.%06ld] ", nanos, rem/1000);
 
+	/* How much buffer space is needed in the worst case */
+	buf_need = sizeof("\\xXX\n");
+
 	for (i = 0, p = msg; i < text_len; i++, p++) {
-		/* 6bytes = "\\x%02x" + '\n' + '\0' */
-		if (bufp - buf >= sizeof(buf) - 6) {
+		if (bufp - buf >= sizeof(buf) - buf_need) {
 			if (write(info->fd_dumpfile, buf, bufp - buf) < 0)
 				return FALSE;
 			bufp = buf;
-- 

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

RE: [PATCH] dump_log_entry: replace magic constant with a variable

[Atsushi Kumagai]

>If the buffer space requirements are changed, the magic constant is
>hard to spot, and the failure is easily overlooked, because it may
>fail only under special conditions.

Good change, I'll push it to devel soon.


Thanks
Atsushi Kumagai

>Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
>---
> makedumpfile.c | 8 +++++---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
>diff --git a/makedumpfile.c b/makedumpfile.c
>index b4b6eca..37dea37 100644
>--- a/makedumpfile.c
>+++ b/makedumpfile.c
>@@ -3949,7 +3949,7 @@ static int
> dump_log_entry(char *logptr, int fp)
> {
> 	char *msg, *p, *bufp;
>-	unsigned int i, text_len;
>+	unsigned int i, text_len, buf_need;
> 	unsigned long long ts_nsec;
> 	char buf[BUFSIZE];
> 	ulonglong nanos;
>@@ -3966,9 +3966,11 @@ dump_log_entry(char *logptr, int fp)
> 	bufp = buf;
> 	bufp += sprintf(buf, "[%5lld.%06ld] ", nanos, rem/1000);
>
>+	/* How much buffer space is needed in the worst case */
>+	buf_need = sizeof("\\xXX\n");
>+
> 	for (i = 0, p = msg; i < text_len; i++, p++) {
>-		/* 6bytes = "\\x%02x" + '\n' + '\0' */
>-		if (bufp - buf >= sizeof(buf) - 6) {
>+		if (bufp - buf >= sizeof(buf) - buf_need) {
> 			if (write(info->fd_dumpfile, buf, bufp - buf) < 0)
> 				return FALSE;
> 			bufp = buf;
>--

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] dump_log_entry: replace magic constant with a variable

[Petr Tesarik]

On Mon, 1 Sep 2014 06:37:10 +0000
Atsushi Kumagai <kumagai-atsushi@mxc.nes.nec.co.jp> wrote:

> >If the buffer space requirements are changed, the magic constant is
> >hard to spot, and the failure is easily overlooked, because it may
> >fail only under special conditions.
> 
> Good change, I'll push it to devel soon.

Thank you!

Petr Tesarik

> Thanks
> Atsushi Kumagai
> 
> >Signed-off-by: Petr Tesarik <ptesarik@suse.cz>
> >---
> > makedumpfile.c | 8 +++++---
> > 1 file changed, 5 insertions(+), 3 deletions(-)
> >
> >diff --git a/makedumpfile.c b/makedumpfile.c
> >index b4b6eca..37dea37 100644
> >--- a/makedumpfile.c
> >+++ b/makedumpfile.c
> >@@ -3949,7 +3949,7 @@ static int
> > dump_log_entry(char *logptr, int fp)
> > {
> > 	char *msg, *p, *bufp;
> >-	unsigned int i, text_len;
> >+	unsigned int i, text_len, buf_need;
> > 	unsigned long long ts_nsec;
> > 	char buf[BUFSIZE];
> > 	ulonglong nanos;
> >@@ -3966,9 +3966,11 @@ dump_log_entry(char *logptr, int fp)
> > 	bufp = buf;
> > 	bufp += sprintf(buf, "[%5lld.%06ld] ", nanos, rem/1000);
> >
> >+	/* How much buffer space is needed in the worst case */
> >+	buf_need = sizeof("\\xXX\n");
> >+
> > 	for (i = 0, p = msg; i < text_len; i++, p++) {
> >-		/* 6bytes = "\\x%02x" + '\n' + '\0' */
> >-		if (bufp - buf >= sizeof(buf) - 6) {
> >+		if (bufp - buf >= sizeof(buf) - buf_need) {
> > 			if (write(info->fd_dumpfile, buf, bufp - buf) < 0)
> > 				return FALSE;
> > 			bufp = buf;
> >--


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec