Re: [PATCH RFC v5 05/16] fuse: make args->in_args[0] to be always the header

[Bernd Schubert <bernd.schubert@fastmail.fm>]

On 11/14/24 22:29, Joanne Koong wrote:
> On Thu, Nov 14, 2024 at 1:05 PM Bernd Schubert
> <bernd.schubert@fastmail.fm> wrote:
>>
>>
>>
>> On 11/14/24 21:57, Joanne Koong wrote:
>>> On Thu, Nov 7, 2024 at 9:04 AM Bernd Schubert <bschubert@ddn.com> wrote:
>>>>
>>>> This change sets up FUSE operations to have headers in args.in_args[0],
>>>> even for opcodes without an actual header. We do this to prepare for
>>>> cleanly separating payload from headers in the future.
>>>>
>>>> For opcodes without a header, we use a zero-sized struct as a
>>>> placeholder. This approach:
>>>> - Keeps things consistent across all FUSE operations
>>>> - Will help with payload alignment later
>>>> - Avoids future issues when header sizes change
>>>>
>>>> Signed-off-by: Bernd Schubert <bschubert@ddn.com>
>>>> ---
>>>>  fs/fuse/dax.c    | 13 ++++++++-----
>>>>  fs/fuse/dev.c    | 24 ++++++++++++++++++++----
>>>>  fs/fuse/dir.c    | 41 +++++++++++++++++++++++++++--------------
>>>>  fs/fuse/fuse_i.h |  7 +++++++
>>>>  fs/fuse/xattr.c  |  9 ++++++---
>>>>  5 files changed, 68 insertions(+), 26 deletions(-)
>>>>
>>>> diff --git a/fs/fuse/dax.c b/fs/fuse/dax.c
>>>> index 12ef91d170bb3091ac35a33d2b9dc38330b00948..e459b8134ccb089f971bebf8da1f7fc5199c1271 100644
>>>> --- a/fs/fuse/dax.c
>>>> +++ b/fs/fuse/dax.c
>>>> @@ -237,14 +237,17 @@ static int fuse_send_removemapping(struct inode *inode,
>>>>         struct fuse_inode *fi = get_fuse_inode(inode);
>>>>         struct fuse_mount *fm = get_fuse_mount(inode);
>>>>         FUSE_ARGS(args);
>>>> +       struct fuse_zero_in zero_arg;
>>>>
>>>>         args.opcode = FUSE_REMOVEMAPPING;
>>>>         args.nodeid = fi->nodeid;
>>>> -       args.in_numargs = 2;
>>>> -       args.in_args[0].size = sizeof(*inargp);
>>>> -       args.in_args[0].value = inargp;
>>>> -       args.in_args[1].size = inargp->count * sizeof(*remove_one);
>>>> -       args.in_args[1].value = remove_one;
>>>> +       args.in_numargs = 3;
>>>> +       args.in_args[0].size = sizeof(zero_arg);
>>>> +       args.in_args[0].value = &zero_arg;
>>>> +       args.in_args[1].size = sizeof(*inargp);
>>>> +       args.in_args[1].value = inargp;
>>>> +       args.in_args[2].size = inargp->count * sizeof(*remove_one);
>>>> +       args.in_args[2].value = remove_one;
>>>>         return fuse_simple_request(fm, &args);
>>>>  }
>>>>
>>>> diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
>>>> index dbc222f9b0f0e590ce3ef83077e6b4cff03cff65..6effef4073da3dad2f6140761eca98147a41d88d 100644
>>>> --- a/fs/fuse/dev.c
>>>> +++ b/fs/fuse/dev.c
>>>> @@ -1007,6 +1007,19 @@ static int fuse_copy_args(struct fuse_copy_state *cs, unsigned numargs,
>>>>
>>>>         for (i = 0; !err && i < numargs; i++)  {
>>>>                 struct fuse_arg *arg = &args[i];
>>>> +
>>>> +               /* zero headers */
>>>> +               if (arg->size == 0) {
>>>> +                       if (WARN_ON_ONCE(i != 0)) {
>>>> +                               if (cs->req)
>>>> +                                       pr_err_once(
>>>> +                                               "fuse: zero size header in opcode %d\n",
>>>> +                                               cs->req->in.h.opcode);
>>>> +                               return -EINVAL;
>>>> +                       }
>>>> +                       continue;
>>>> +               }
>>>> +
>>>>                 if (i == numargs - 1 && argpages)
>>>>                         err = fuse_copy_pages(cs, arg->size, zeroing);
>>>>                 else
>>>> @@ -1662,6 +1675,7 @@ static int fuse_retrieve(struct fuse_mount *fm, struct inode *inode,
>>>>         size_t args_size = sizeof(*ra);
>>>>         struct fuse_args_pages *ap;
>>>>         struct fuse_args *args;
>>>> +       struct fuse_zero_in zero_arg;
>>>>
>>>>         offset = outarg->offset & ~PAGE_MASK;
>>>>         file_size = i_size_read(inode);
>>>> @@ -1688,7 +1702,7 @@ static int fuse_retrieve(struct fuse_mount *fm, struct inode *inode,
>>>>         args = &ap->args;
>>>>         args->nodeid = outarg->nodeid;
>>>>         args->opcode = FUSE_NOTIFY_REPLY;
>>>> -       args->in_numargs = 2;
>>>> +       args->in_numargs = 3;
>>>>         args->in_pages = true;
>>>>         args->end = fuse_retrieve_end;
>>>>
>>>> @@ -1715,9 +1729,11 @@ static int fuse_retrieve(struct fuse_mount *fm, struct inode *inode,
>>>>         }
>>>>         ra->inarg.offset = outarg->offset;
>>>>         ra->inarg.size = total_len;
>>>> -       args->in_args[0].size = sizeof(ra->inarg);
>>>> -       args->in_args[0].value = &ra->inarg;
>>>> -       args->in_args[1].size = total_len;
>>>> +       args->in_args[0].size = sizeof(zero_arg);
>>>> +       args->in_args[0].value = &zero_arg;
>>>> +       args->in_args[1].size = sizeof(ra->inarg);
>>>> +       args->in_args[1].value = &ra->inarg;
>>>> +       args->in_args[2].size = total_len;
>>>>
>>>>         err = fuse_simple_notify_reply(fm, args, outarg->notify_unique);
>>>>         if (err)
>>>
>>> Do we also need to add a zero arg header for FUSE_READLINK,
>>> FUSE_DESTROY, and FUSE_BATCH_FORGET requests as well?
>>>
>>
>> Thanks for looking at the patch! I should have added to the commit message
>> that I didn't modify these, as they don't have an in argument at all.
>>
> 
> Thanks for clarifying! (and apologies for the late review. I haven't
> been keeping up with these patches since RFC v3 but I'm planning to
> get up to speed and take a deeper look at these tomorrow + next week).

No worries at all... I'm also very late with reviewing your patches. 
I'm close for the next fuse-io-version, just fixing some bg accounting
issues that had been in all rfc versions so far.

> 
> I think the FUSE_BATCH_FORGET request does use in args, depending on
> the number of forget requests.

Ah right, but it does not use fuse_copy_args and args->in_args[idx] - 
is very special. And just looking it up again, the header is in the
right place. Issue would be more for over-io-uring to copy into the
payload. However, current over-io-uring patches don't handle forgets
at all - it goes over /dev/fuse. Unless you disagree, I think we can
do forgets later on over io-uring as optimization.


Thanks,
Bernd