[PATCH] fix: nvme_update_ns_info method should be called even if nvme_ms_ids_equal return false

Chaitanya Kulkarni chaitanyak at nvidia.com
Tue May 24 13:54:15 PDT 2022


>> Please try this:
>>
>>
>> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
>> index 5a98a7de09642..cb4adc0c22843 100644
>> --- a/drivers/nvme/host/pci.c
>> +++ b/drivers/nvme/host/pci.c
>> @@ -3453,6 +3453,8 @@ static const struct pci_device_id
>> nvme_id_table[] = {
>>                  .driver_data = NVME_QUIRK_NO_DEEPEST_PS, },
>>          { PCI_DEVICE(0x2646, 0x2263),   /* KINGSTON A2000 NVMe SSD
>> */
>>                  .driver_data = NVME_QUIRK_NO_DEEPEST_PS, },
>> +       { PCI_DEVICE(0x1e4B, 0x1001),   /* MAXIO MAP1001 */
>> +               .driver_data = NVME_QUIRK_BOGUS_NID, },
>>          { PCI_DEVICE(0x1e4B, 0x1002),   /* MAXIO MAP1002 */
>>                  .driver_data = NVME_QUIRK_BOGUS_NID, },
>>          { PCI_DEVICE(0x1e4B, 0x1202),   /* MAXIO MAP1202 */
>>
>>
> I applied the patch against kernel 5.18.0
> As dmesg log below shows, the quirk is applied correctly and resume
> from suspend is working as expected.
> Thank you
> 
> [    1.684064] nvme 0000:04:00.0: platform quirk: setting simple
> suspend
> [    1.684119] nvme nvme0: pci function 0000:04:00.0
> [    1.696073] nvme nvme0: missing or invalid SUBNQN field.
> [    1.699332] nvme nvme0: allocated 16 MiB host memory buffer.
> [    1.701789] nvme nvme0: 8/0/0 default/read/poll queues
> [    1.707999] nvme nvme0: Ignoring bogus Namespace Identifiers
> [    1.710368]  nvme0n1: p1 p2 p3 p4 p5 p6 p7
> [   36.560018] XFS (nvme0n1p5): Mounting V5 Filesystem
> [   36.982002] XFS (nvme0n1p5): Ending clean mount
> [   91.027799] nvme nvme0: 8/0/0 default/read/poll queues
> [   91.034207] nvme nvme0: Ignoring bogus Namespace Identifiers


Feel free to add this if you are planning to send as a formal
patch.

Reviewed-by: Chaitanya Kulkarni <kch at nvidia.com>

-ck




More information about the Linux-nvme mailing list