nvme: cleanup double shift issue (77016199) · Commits · 戴 / test

The problem here is that set_bit() and test_bit() take a bit number so we should be passing 0 but instead we're passing (1 << 0) which leads to a double shift. It doesn't cause a runtime bug in the current code because it's done consistently and we only set that one bit. I decided to just re-use NVME_AER_NOTICE_NS_CHANGED instead of introducing a new define for this. Signed-off-by:

Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by:

Sagi Grimberg <sagi@grimberg.me> Signed-off-by:

Christoph Hellwig <hch@lst.de> Signed-off-by:

Jens Axboe <axboe@kernel.dk>

drivers/nvme/host/core.c

+2 −2

Original line number	Diff line number	Diff line
		@@ -3245,7 +3245,7 @@ static void nvme_scan_work(struct work_struct *work)

		WARN_ON_ONCE(!ctrl->tagset);

		if (test_and_clear_bit(EVENT_NS_CHANGED, &ctrl->events)) {
		if (test_and_clear_bit(NVME_AER_NOTICE_NS_CHANGED, &ctrl->events)) {
		if (nvme_scan_changed_ns_log(ctrl))
		goto out_sort_namespaces;
		dev_info(ctrl->device, "rescanning namespaces.\n");
		@@ -3386,7 +3386,7 @@ static void nvme_handle_aen_notice(struct nvme_ctrl *ctrl, u32 result)
		{
		switch ((result & 0xff00) >> 8) {
		case NVME_AER_NOTICE_NS_CHANGED:
		set_bit(EVENT_NS_CHANGED, &ctrl->events);
		set_bit(NVME_AER_NOTICE_NS_CHANGED, &ctrl->events);
		nvme_queue_scan(ctrl);
		break;
		case NVME_AER_NOTICE_FW_ACT_STARTING:

drivers/nvme/host/nvme.h

+0 −1

Original line number	Diff line number	Diff line
		@@ -194,7 +194,6 @@ struct nvme_ctrl {
		struct delayed_work ka_work;
		struct nvme_command ka_cmd;
		struct work_struct fw_act_work;
		#define EVENT_NS_CHANGED (1 << 0)
		unsigned long events;

		/* Power saving configuration */

Admin message