summaryrefslogtreecommitdiff
path: root/fuzz.c
diff options
context:
space:
mode:
authordobin <dobin@broken.ch>2018-04-14 15:59:46 +0200
committerdobin <dobin@broken.ch>2018-04-14 15:59:46 +0200
commitea3ae442d4c23ebf2e761ccba1f3422eda8e25fc (patch)
tree59f6bc77bf50fb78e43d55b5f0183f3982bf5de8 /fuzz.c
parent040a98ed1359b125daec314729378bdc900f603b (diff)
downloadhonggfuzz-ea3ae442d4c23ebf2e761ccba1f3422eda8e25fc.tar.gz
remove second reapchild in socketfuzzer
The reapchild() is already performed via subproc_Run(). This second reapchild() makes honggfuzz timeout when waiting for SIGIO, because we send the signal only once. It is basically unecessary.
Diffstat (limited to 'fuzz.c')
-rw-r--r--fuzz.c11
1 files changed, 0 insertions, 11 deletions
diff --git a/fuzz.c b/fuzz.c
index 0a010cce..f9e7b391 100644
--- a/fuzz.c
+++ b/fuzz.c
@@ -513,17 +513,6 @@ static void fuzz_fuzzLoopSocket(run_t* run) {
}
report_Report(run);
-
- /* Try to identify if the target crashed.
- This information will be used in the next iteration of
- this loop, to restart the target if necessary.
- The fuzzer will be also notified.
-
- Crash identification does not need to work 100%, as the external fuzzer
- can also detect timeouts on the target server, and will
- notify us. */
- LOG_D("------[ 4: reap child");
- arch_reapChild(run);
}
static void* fuzz_threadNew(void* arg) {