[PATCHv2] nvme-pci: Fix EEH failure on ppc after subsystem reset

[Nilay Shroff]

On 4/8/24 15:56, Nilay Shroff wrote:
> If the nvme subsyetm reset causes the loss of communication to the nvme
> adapter then EEH could potnetially recover the adapter. The detection of
> comminication loss to the adapter only happens when the nvme driver
> attempts to read an MMIO register.
> 
> The nvme subsystem reset command writes 0x4E564D65 to NSSR register and
> schedule adapter reset.In the case nvme subsystem reset caused the loss
> of communication to the nvme adapter then either IO timeout event or
> adapter reset handler could detect it. If IO timeout event could detect
> loss of communication then EEH handler is able to recover the communication
> to the adapter. This change was implemented in commit 651438bb0af5
> ("nvme-pci: Fix EEH failure on ppc"). However if the adapter communication
> loss is detected during nvme reset work then EEH is unable to successfully
> finish the adapter recovery.
> 
> This patch ensures that,
> - nvme reset work can observer pci channel is offline (at-least on the
>   paltfrom which supports EEH recovery) after a failed MMIO read and
>   contains reset work forward progress and marking controller state to
>   DEAD. Thus we give a fair chance to EEH handler to recover the nvme
>   adapter.
> 
> - if pci channel "frozen" error is detected while controller is already
>   in the RESETTING state then don't try (re-)setting controller state to
>   RESETTING which would otherwise obviously fail and we may prematurely
>   breaks out of the EEH recovery handling.
> 
> - if pci channel "frozen" error is detected while reset work is in progress
>   then wait until reset work is finished before proceeding with nvme dev
>   disable. This would ensure that the reset work doesn't race with the
>   pci error handler code and both error handler and reset work forward
>   progress without blocking.
> 
> Signed-off-by: Nilay Shroff <nilay at linux.ibm.com>
> ---
> Changes from v1:
>   - Allow a controller to reset from a connecting state (Keith)
> 
>   - Fix race condition between reset work and pci error handler 
>     code which may contain reset work and EEH recovery from 
>     forward progress (Keith)
> 
>  drivers/nvme/host/core.c |  1 +
>  drivers/nvme/host/pci.c  | 19 ++++++++++++++++---
>  2 files changed, 17 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
> index 27281a9a8951..b3fe1a02c171 100644
> --- a/drivers/nvme/host/core.c
> +++ b/drivers/nvme/host/core.c
> @@ -557,6 +557,7 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
>  		switch (old_state) {
>  		case NVME_CTRL_NEW:
>  		case NVME_CTRL_LIVE:
> +		case NVME_CTRL_CONNECTING:
>  			changed = true;
>  			fallthrough;
>  		default:
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index 8e0bb9692685..553bf0ec5f5c 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -2776,6 +2776,16 @@ static void nvme_reset_work(struct work_struct *work)
>   out_unlock:
>  	mutex_unlock(&dev->shutdown_lock);
>   out:
> +	/*
> +	 * If PCI recovery is ongoing then let it finish first
> +	 */
> +	if (pci_channel_offline(to_pci_dev(dev->dev))) {
> +		if (nvme_ctrl_state(&dev->ctrl) == NVME_CTRL_RESETTING ||
> +			nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_RESETTING)) {
> +			dev_warn(dev->ctrl.device, "Let pci error recovery finish!\n");
> +			return;
> +		}
> +	}
>  	/*
>  	 * Set state to deleting now to avoid blocking nvme_wait_reset(), which
>  	 * may be holding this pci_dev's device lock.
> @@ -3295,10 +3305,13 @@ static pci_ers_result_t nvme_error_detected(struct pci_dev *pdev,
>  	case pci_channel_io_frozen:
>  		dev_warn(dev->ctrl.device,
>  			"frozen state error detected, reset controller\n");
> -		if (!nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_RESETTING)) {
> -			nvme_dev_disable(dev, true);
> -			return PCI_ERS_RESULT_DISCONNECT;
> +		if (nvme_ctrl_state(&dev->ctrl) != NVME_CTRL_RESETTING) {
> +			if (!nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_RESETTING)) {
> +				nvme_dev_disable(dev, true);
> +				return PCI_ERS_RESULT_DISCONNECT;
> +			}
>  		}
> +		flush_work(&dev->ctrl.reset_work);
>  		nvme_dev_disable(dev, false);
>  		return PCI_ERS_RESULT_NEED_RESET;
>  	case pci_channel_io_perm_failure:
A gentle ping... Can I get some feedback?
For reference, link to first version of the patch is here: https://lore.kernel.org/all/20240209050342.406184-1-nilay@linux.ibm.com/

Thanks,
--Nilay