On 2/15/24 3:37 PM, David Wei wrote:
> On 2024-02-15 15:32, Jens Axboe wrote:
>> 1 usec is not as short as it used to be, and it makes sense to allow 0
>> for a busy poll timeout - this means just do one loop to check if we
>> have anything available. Add a separate ->napi_enabled to check if napi
>> has been enabled or not.
>>
>> While at it, move the writing of the ctx napi values after we've copied
>> the old values back to userspace. This ensures that if the call fails,
>> we'll be in the same state as we were before, rather than some
>> indeterminate state.
>>
>> Signed-off-by: Jens Axboe <axboe@xxxxxxxxx>
>>
>> ---
>>
>> diff --git a/include/linux/io_uring_types.h b/include/linux/io_uring_types.h
>> index 4fe7af8a4907..bd7071aeec5d 100644
>> --- a/include/linux/io_uring_types.h
>> +++ b/include/linux/io_uring_types.h
>> @@ -420,6 +420,7 @@ struct io_ring_ctx {
>> /* napi busy poll default timeout */
>> unsigned int napi_busy_poll_to;
>> bool napi_prefer_busy_poll;
>> + bool napi_enabled;
>>
>> DECLARE_HASHTABLE(napi_ht, 4);
>> #endif
>> diff --git a/io_uring/napi.c b/io_uring/napi.c
>> index b234adda7dfd..e653927a376e 100644
>> --- a/io_uring/napi.c
>> +++ b/io_uring/napi.c
>> @@ -227,12 +227,12 @@ int io_register_napi(struct io_ring_ctx *ctx, void __user *arg)
>> if (napi.pad[0] || napi.pad[1] || napi.pad[2] || napi.resv)
>> return -EINVAL;
>>
>> - WRITE_ONCE(ctx->napi_busy_poll_to, napi.busy_poll_to);
>> - WRITE_ONCE(ctx->napi_prefer_busy_poll, !!napi.prefer_busy_poll);
>> -
>> if (copy_to_user(arg, &curr, sizeof(curr)))
>> return -EFAULT;
>>
>> + WRITE_ONCE(ctx->napi_busy_poll_to, napi.busy_poll_to);
>> + WRITE_ONCE(ctx->napi_prefer_busy_poll, !!napi.prefer_busy_poll);
>> + WRITE_ONCE(ctx->napi_enabled, true);
>> return 0;
>> }
>>
>> @@ -256,6 +256,7 @@ int io_unregister_napi(struct io_ring_ctx *ctx, void __user *arg)
>>
>> WRITE_ONCE(ctx->napi_busy_poll_to, 0);
>> WRITE_ONCE(ctx->napi_prefer_busy_poll, false);
>> + WRITE_ONCE(ctx->napi_enabled, true);
>
> Should this be false?
It should indeed... Updated, thanks!
--
Jens Axboe
