Closer examination shows that nvme and CAM both normally zero-fill

allocations (for req and ccb, which ultimately contain the
nvme_cmd). As such, we can micro-optimize these routines. Add a
comment to this effect, and bzero the ccb used to make the requests
for the nda dump rotuine so it more closely matches a ccb allocated
with xpt_get_ccb().

Sponsored by: Netflix

1 files changed, 2 insertions, 0 deletions

diff --git a/sys/cam/nvme/nvme_da.c b/sys/cam/nvme/nvme_da.c
index 807cf3e0ea15..f0f438b4fb79 100644
--- a/sys/cam/nvme/nvme_da.c
+++ b/sys/cam/nvme/nvme_da.c
@@ -398,6 +398,8 @@ ndadump(void *arg, void *virtual, vm_offset_t physical, off_t offset, size_t len
 		return (ENXIO);
 	}
 
+	/* xpt_get_ccb returns a zero'd allocation for the ccb, mimic that here */
+	memset(&nvmeio, 0, sizeof(nvmeio));
 	if (length > 0) {
 		xpt_setup_ccb(&nvmeio.ccb_h, periph->path, CAM_PRIORITY_NORMAL);
 		nvmeio.ccb_h.ccb_state = NDA_CCB_DUMP;