Re: [PATCH v2 1/5] NFSD: OFFLOAD_CANCEL should mark an async COPY as completed

[Chuck Lever <chuck.lever@oracle.com>]

On 3/2/25 4:35 PM, Jeff Layton wrote:
> On Sat, 2025-03-01 at 13:31 -0500, cel@kernel.org wrote:
>> From: Chuck Lever <chuck.lever@oracle.com>
>>
>> Update the status of an async COPY operation when it has been
>> stopped. OFFLOAD_STATUS needs to indicate that the COPY is no longer
>> running.
>>
>> Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
>> ---
>>  fs/nfsd/nfs4proc.c | 5 ++++-
>>  1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
>> index f6e06c779d09..9a0e68aa246f 100644
>> --- a/fs/nfsd/nfs4proc.c
>> +++ b/fs/nfsd/nfs4proc.c
>> @@ -1379,8 +1379,11 @@ static void nfs4_put_copy(struct nfsd4_copy *copy)
>>  static void nfsd4_stop_copy(struct nfsd4_copy *copy)
>>  {
>>  	trace_nfsd_copy_async_cancel(copy);
>> -	if (!test_and_set_bit(NFSD4_COPY_F_STOPPED, &copy->cp_flags))
>> +	if (!test_and_set_bit(NFSD4_COPY_F_STOPPED, &copy->cp_flags)) {
>>  		kthread_stop(copy->copy_task);
>> +		copy->nfserr = nfs_ok;
>> +		set_bit(NFSD4_COPY_F_COMPLETED, &copy->cp_flags);
>> +	}
>>  	nfs4_put_copy(copy);
>>  }
>>  
> 
> STOPPED and COMPLETED seem to basically track each other. What's the
> distinction between the two bits? When is one set, but not the other?

IIRC:

COMPLETED means the background COPY is no longer running. It prevents
kthread_stop() from being called more than once on the same kthread.

STOPPED means that the background COPY is no longer running because
something stopped it before it could finish normally -- more or less
this is like being signaled.

We could name it "ABEND" or something else.

-- 
Chuck Lever