[libvirt] [PATCH] qemu: fix text block info parsing

Detected by Coverity. p (the pointer to the string) is always true; when in reality, we wanted to know whether the integer value of the just-parsed string is '0' or '1'. Logic bug since commit b1b5b51. * src/qemu/qemu_monitor_text.c (qemuMonitorTextGetBlockInfo): Set results to proper value. --- Second bug introduced by the same commit. And I think I reviewed it originally. Sorry for not catching it earlier. src/qemu/qemu_monitor_text.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index 1eb9846..49c469b 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -817,19 +817,19 @@ int qemuMonitorTextGetBlockInfo(qemuMonitorPtr mon, if (virStrToLong_i(p, &dummy, 10, &tmp) == -1) VIR_DEBUG("error reading removable: %s", p); else - info->removable = p != NULL; + info->removable = tmp != 0; } else if (STRPREFIX(p, "locked=")) { p += strlen("locked="); if (virStrToLong_i(p, &dummy, 10, &tmp) == -1) VIR_DEBUG("error reading locked: %s", p); else - info->locked = p ? true : false; + info->locked = tmp != 0; } else if (STRPREFIX(p, "tray_open=")) { p += strlen("tray_open="); if (virStrToLong_i(p, &dummy, 10, &tmp) == -1) VIR_DEBUG("error reading tray_open: %s", p); else - info->tray_open = p ? true : false; + info->tray_open = tmp != 0; } else { /* ignore because we don't parse all options */ } -- 1.7.4.4

On Wed, Oct 12, 2011 at 06:30:08PM -0600, Eric Blake wrote:
Detected by Coverity. p (the pointer to the string) is always true; when in reality, we wanted to know whether the integer value of the just-parsed string is '0' or '1'. Logic bug since commit b1b5b51.
* src/qemu/qemu_monitor_text.c (qemuMonitorTextGetBlockInfo): Set results to proper value. ---
Second bug introduced by the same commit. And I think I reviewed it originally. Sorry for not catching it earlier.
Absolutely no need to be sorry, 1/ you're the one ACK'ing the most patches 2/ I think you filter out bugs way better than me for example statistics will play against you, but nobody will complain !
src/qemu/qemu_monitor_text.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index 1eb9846..49c469b 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -817,19 +817,19 @@ int qemuMonitorTextGetBlockInfo(qemuMonitorPtr mon, if (virStrToLong_i(p, &dummy, 10, &tmp) == -1) VIR_DEBUG("error reading removable: %s", p); else - info->removable = p != NULL; + info->removable = tmp != 0;
I would be tempted to put braces info->removable = (tmp != 0); to make it obvious that the second part is computation, not cascaded affectations
} else if (STRPREFIX(p, "locked=")) { p += strlen("locked="); if (virStrToLong_i(p, &dummy, 10, &tmp) == -1) VIR_DEBUG("error reading locked: %s", p); else - info->locked = p ? true : false; + info->locked = tmp != 0; } else if (STRPREFIX(p, "tray_open=")) { p += strlen("tray_open="); if (virStrToLong_i(p, &dummy, 10, &tmp) == -1) VIR_DEBUG("error reading tray_open: %s", p); else - info->tray_open = p ? true : false; + info->tray_open = tmp != 0; } else { /* ignore because we don't parse all options */ }
But it's cosmetic, ACK Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

On 10/13/2011 03:15 AM, Daniel Veillard wrote:
On Wed, Oct 12, 2011 at 06:30:08PM -0600, Eric Blake wrote:
Detected by Coverity. p (the pointer to the string) is always true; when in reality, we wanted to know whether the integer value of the just-parsed string is '0' or '1'. Logic bug since commit b1b5b51.
* src/qemu/qemu_monitor_text.c (qemuMonitorTextGetBlockInfo): Set results to proper value. ---
- info->removable = p != NULL; + info->removable = tmp != 0;
I would be tempted to put braces info->removable = (tmp != 0); to make it obvious that the second part is computation, not cascaded affectations
Done,
But it's cosmetic, ACK
and pushed. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
participants (2)
-
Daniel Veillard
-
Eric Blake