Re: [PATCH v4 08/15] nvme: Implement cross-controller reset recovery

[Mohamed Khalfella <mkhalfella@purestorage.com>]

On Tue 2026-04-07 07:39:09 +0200, Hannes Reinecke wrote:
> On 3/31/26 18:47, Mohamed Khalfella wrote:
> > On Mon 2026-03-30 12:50:24 +0200, Hannes Reinecke wrote:
> >> On 3/28/26 01:43, Mohamed Khalfella wrote:
> >>> A host that has more than one path connecting to an nvme subsystem
> >>> typically has an nvme controller associated with every path. This is
> >>> mostly applicable to nvmeof. If one path goes down, inflight IOs on that
> >>> path should not be retried immediately on another path because this
> >>> could lead to data corruption as described in TP4129. TP8028 defines
> >>> cross-controller reset mechanism that can be used by host to terminate
> >>> IOs on the failed path using one of the remaining healthy paths. Only
> >>> after IOs are terminated, or long enough time passes as defined by
> >>> TP4129, inflight IOs should be retried on another path. Implement core
> >>> cross-controller reset shared logic to be used by the transports.
> >>>
> >>> Signed-off-by: Mohamed Khalfella <mkhalfella@purestorage.com>
> >>> ---
> >>>    drivers/nvme/host/constants.c |   1 +
> >>>    drivers/nvme/host/core.c      | 145 ++++++++++++++++++++++++++++++++++
> >>>    drivers/nvme/host/nvme.h      |   9 +++
> >>>    3 files changed, 155 insertions(+)
> >>>
> [ .. ]
> >>> +
> >>> +int nvme_fence_ctrl(struct nvme_ctrl *ictrl)
> >>> +{
> >>> +	unsigned long deadline, timeout;
> >>> +	struct nvme_ctrl *sctrl;
> >>> +	u32 min_cntlid = 0;
> >>> +	int ret;
> >>> +
> >>> +	timeout = nvme_fence_timeout_ms(ictrl);
> >>> +	dev_info(ictrl->device, "attempting CCR, timeout %lums\n", timeout);
> >>> +
> >>> +	deadline = jiffies + msecs_to_jiffies(timeout);
> >>> +	while (time_is_after_jiffies(deadline)) {
> >>> +		sctrl = nvme_find_ctrl_ccr(ictrl, min_cntlid);
> >>> +		if (!sctrl) {
> >>> +			dev_dbg(ictrl->device,
> >>> +				"failed to find source controller\n");
> >>> +			return -EIO;
> >>> +		}
> >>> +
> >>> +		ret = nvme_issue_wait_ccr(sctrl, ictrl, deadline);
> >>> +		if (!ret) {
> >>> +			dev_info(ictrl->device, "CCR succeeded using %s\n",
> >>> +				 dev_name(sctrl->device));
> >>> +			nvme_put_ctrl_ccr(sctrl);
> >>> +			return 0;
> >>> +		}
> >>> +
> >>> +		min_cntlid = sctrl->cntlid + 1;
> >>> +		nvme_put_ctrl_ccr(sctrl);
> >>> +
> >>> +		if (ret == -EIO) /* CCR command failed */
> >>> +			continue;
> >>> +
> >>> +		/* CCR operation failed or timed out */
> >>> +		return ret;
> >>> +	}
> >>> +
> >>> +	dev_info(ictrl->device, "CCR operation timeout\n");
> >>> +	return -ETIMEDOUT;
> >>> +}
> >>
> >> Please restructure the loop.
> >> Having a comment 'CCR operation failed or timed out',
> >> returning a status, and then have a comment
> >> 'CCR operation timeout' _after_ the return is confusing.
> > 
> > I can change /* CCR operation failed or timed out */ to something like
> > 
> > /*
> >   * Source controller accepted CCR command but CCR operation
> >   * timed out or failed. Retrying another path is not likely
> >   * to succeed, return an error.
> >   */
> > 
> > And change the log line "CCR operation timeout\n" outside the while
> > loop to "fencing timedout\n".
> > 
> > Will this help?
> > 
> Yes, thank you.

Will do that.

> 
> Cheers,
> 
> Hannes
> -- 
> Dr. Hannes Reinecke                  Kernel Storage Architect
> hare@suse.de                                +49 911 74053 688
> SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
> HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich