On 02/01/2012 04:17 PM, Eric Blake wrote:
> @@ -516,7 +520,6 @@ networkBuildDnsmasqArgv(virNetworkObjPtr
network,
> int i;
>
> for (i = 0; i < dns->ntxtrecords; i++) {
> - char *record = NULL;
> if (virAsprintf(&record, "%s,%s",
> dns->txtrecords[i].name,
> dns->txtrecords[i].value) < 0) {
Hmm, we can simplify this loop even further, by using the right function
in the first place.
ACK to what you had, and I'm squashing this in then pushing:
Serves me right for testing, then modifying, then pushing without retesting.
- virCommandAddArgPair(cmd, "--txt-record",
record);
- VIR_FREE(record);
+ virCommandAddArg(cmd, "--txt-record");
+ virCommandAddArgFormat(cmd, "%s,%s",
+ dns->txtrecords[i].name,
+ dns->txtrecords[i].value);
ArgPair adds an equal sign. I'm pushing this followup under the
build-breaker rule (since 'make check' indeed catches the otherwise
harmless change in the generated command line).
From 0aaf88e800673b0fc204458588af6a96631a81c0 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake(a)redhat.com>
Date: Wed, 1 Feb 2012 16:42:33 -0700
Subject: [PATCH] network: fix testsuite regression
I slightly botched commit be9fb5a - I converted '--arg=value' to
'--arg value', which has no semantic change, but did trip up the
testsuite.
* src/network/bridge_driver.c (networkBuildDnsmasqArgv): Restore
expected output.
---
src/network/bridge_driver.c | 3 +--
1 files changed, 1 insertions(+), 2 deletions(-)
diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c
index 220f99e..57ebb9f 100644
--- a/src/network/bridge_driver.c
+++ b/src/network/bridge_driver.c
@@ -520,8 +520,7 @@ networkBuildDnsmasqArgv(virNetworkObjPtr network,
int i;
for (i = 0; i < dns->ntxtrecords; i++) {
- virCommandAddArg(cmd, "--txt-record");
- virCommandAddArgFormat(cmd, "%s,%s",
+ virCommandAddArgFormat(cmd, "--txt-record=%s,%s",
dns->txtrecords[i].name,
dns->txtrecords[i].value);
}
--
1.7.7.6
--
Eric Blake eblake(a)redhat.com +1-919-301-3266
Libvirt virtualization library
http://libvirt.org