summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorMartin Roth <martinroth@chromium.org>2016-08-15 14:55:37 -0600
committerchrome-bot <chrome-bot@chromium.org>2016-08-22 23:27:18 -0700
commit56cf05339f388a4d8de707cfcc77bf3230ccd715 (patch)
treecea9a2b0984b077f3c82ac27835a61345d39efe1
parent61c45fb33e9ea1b1db6e195c093b02444d303cce (diff)
downloadchrome-ec-56cf05339f388a4d8de707cfcc77bf3230ccd715.tar.gz
chip/g/sps: Initialize received_data pointer to NULL
I don't think that the received_data pointer in sps_rx_interrupt can actually be used without initialization in the current code flow, but GCC 5.3 thinks that it can, and throws an error. sps_check_rx() always sets the pointer or returns 0. In either case we should be ok, but initializing it to null here won't hurt anything since it's not masking an actual error. BUG=None BRANCH=None TEST=Build all boards successfully. Change-Id: Idaea7efd1af3067c492a3ade4c89104e975f4c1d Signed-off-by: Martin Roth <martinroth@chromium.org> Reviewed-on: https://chromium-review.googlesource.com/370663 Reviewed-by: David Hendricks <dhendrix@chromium.org> Reviewed-by: Mary Ruthven <mruthven@chromium.org>
-rw-r--r--chip/g/sps.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/chip/g/sps.c b/chip/g/sps.c
index 0ce1d6be41..923e39398c 100644
--- a/chip/g/sps.c
+++ b/chip/g/sps.c
@@ -315,7 +315,7 @@ static void sps_advance_rx(int port, int data_size)
static void sps_rx_interrupt(uint32_t port, int cs_deasserted)
{
for (;;) {
- uint8_t *received_data;
+ uint8_t *received_data = NULL;
size_t data_size;
data_size = sps_check_rx(port, &received_data);