Re: [RFC PATCH] usb: storage: uas: limit consecutive device resets in error handling

[Greg KH]

On Wed, Jul 01, 2026 at 01:03:21PM +0900, Sergey Senozhatsky wrote:
> When a UAS storage device experiences persistent wire or hardware IO
> failures, commands time out and the SCSI error handler thread invokes
> uas_eh_device_reset_handler().  If usb_reset_device() succeeds at the
> USB hub level but the underlying drive remains unresponsive, the reset
> handler returns SUCCESS. SCSI EH then requeues pending commands with
> DID_RESET (ACTION_RETRY), causing them to time out again 30 seconds
> later in an infinite loop.  This blocks block layer queues indefinitely:
> 
> [..]
>  sd 0:0:0:0: [sda] tag#4 uas_eh_abort_handler 0 uas-tag 1 inflight: CMD
>  sd 0:0:0:0: [sda] tag#4 CDB: Write(10) 2a 00 00 d3 98 08 00 04 00 00
>  sd 0:0:0:0: [sda] tag#0 uas_eh_abort_handler 0 uas-tag 2 inflight: CMD OUT
>  sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 00 00 d3 9c 08 00 04 00 00
>  scsi host0: uas_eh_device_reset_handler start
>  usb 2-1.3: reset SuperSpeed Plus Gen 2x1 USB device number 4 using xhci_hcd
>  scsi host0: uas_eh_device_reset_handler success
>  sd 0:0:0:0: [sda] tag#3 uas_eh_abort_handler 0 uas-tag 3 inflight: CMD IN
>  sd 0:0:0:0: [sda] tag#3 CDB: Read(10) 28 00 00 00 00 00 00 00 20 00
>  scsi host0: uas_eh_device_reset_handler start
>  sd 0:0:0:0: [sda] tag#1 uas_zap_pending 0 uas-tag 1 inflight: CMD
>  sd 0:0:0:0: [sda] tag#1 CDB: Write(10) 2a 00 00 d3 98 08 00 04 00 00
>  sd 0:0:0:0: [sda] tag#2 uas_zap_pending 0 uas-tag 2 inflight: CMD
>  sd 0:0:0:0: [sda] tag#2 CDB: Write(10) 2a 00 00 d3 9c 08 00 04 00 00
>  usb 2-1.3: reset SuperSpeed Plus Gen 2x1 USB device number 4 using xhci_hcd
>  scsi host0: uas_eh_device_reset_handler success
> [..]
> 
> Introduce a runtime-configurable module parameter 'reset_limit' (default
> 3) and track consecutive resets in devinfo->reset_cnt.  When a productive
> block layer command completes successfully (SUBMITTED_BY_BLOCK_LAYER),
> reset the counter to zero.  If consecutive resets exceed reset_limit,
> abort the loop by completing pending commands with DID_NO_CONNECT and
> returning FAILED.  This allows SCSI EH to offline the unresponsive
> device.
> 
> Signed-off-by: Sergey Senozhatsky <senozhatsky@xxxxxxxxxxxx>
> ---
>  drivers/usb/storage/uas.c | 22 ++++++++++++++++++++++
>  1 file changed, 22 insertions(+)
> 
> diff --git a/drivers/usb/storage/uas.c b/drivers/usb/storage/uas.c
> index 265162981269..a63c66c8bbad 100644
> --- a/drivers/usb/storage/uas.c
> +++ b/drivers/usb/storage/uas.c
> @@ -32,6 +32,10 @@
>  
>  #define MAX_CMNDS 256
>  
> +static int uas_reset_limit = 3;
> +module_param_named(reset_limit, uas_reset_limit, int, 0644);
> +MODULE_PARM_DESC(reset_limit, "Maximum number of consecutive device resets during error handling before failing");

This is not the 1990's, we do not add module parameters for issues that
should be properly solved either automatically, or on a per-device
basis.

There's no way that ChromeOs wants to attempt to track this module
parameter as a bootline config option, right?

thanks,

greg k-h