On Tue, Jun 26, 2018 at 4:09 AM cgxu519 <cgxu519@xxxxxxx> wrote:
>
>
>
> On 06/24/2018 05:05 PM, Ilya Dryomov wrote:
> > On Sun, Jun 24, 2018 at 10:57 AM cgxu519 <cgxu519@xxxxxxx> wrote:
> >>
> >> On 06/24/2018 04:49 PM, Ilya Dryomov wrote:
> >>> On Sat, Jun 23, 2018 at 2:55 PM Chengguang Xu <cgxu519@xxxxxxx> wrote:
> >>>> Though val_size1/val_size2 are both type sizt_t but
> >>>> ceph_pagelist_encode_string() can only handle string
> >>>> size max to U32_MAX, so change hardcode 8 to sizeof(u32).
> >>>>
> >>>> Signed-off-by: Chengguang Xu <cgxu519@xxxxxxx>
> >>>> ---
> >>>> fs/ceph/acl.c | 6 ++++--
> >>>> 1 file changed, 4 insertions(+), 2 deletions(-)
> >>>>
> >>>> diff --git a/fs/ceph/acl.c b/fs/ceph/acl.c
> >>>> index 3351ea14390b..f736108604a6 100644
> >>>> --- a/fs/ceph/acl.c
> >>>> +++ b/fs/ceph/acl.c
> >>>> @@ -219,7 +219,8 @@ int ceph_pre_init_acls(struct inode *dir, umode_t *mode,
> >>>>
> >>>> if (acl) {
> >>>> size_t len = strlen(XATTR_NAME_POSIX_ACL_ACCESS);
> >>>> - err = ceph_pagelist_reserve(pagelist, len + val_size1 + 8);
> >>>> + err = ceph_pagelist_reserve(pagelist, len + val_size1 +
> >>>> + sizeof(u32));
> >>>> if (err)
> >>>> goto out_err;
> >>>> ceph_pagelist_encode_string(pagelist, XATTR_NAME_POSIX_ACL_ACCESS,
> >>>> @@ -233,7 +234,8 @@ int ceph_pre_init_acls(struct inode *dir, umode_t *mode,
> >>>> }
> >>>> if (default_acl) {
> >>>> size_t len = strlen(XATTR_NAME_POSIX_ACL_DEFAULT);
> >>>> - err = ceph_pagelist_reserve(pagelist, len + val_size2 + 8);
> >>>> + err = ceph_pagelist_reserve(pagelist, len + val_size2 +
> >>>> + sizeof(u32));
> >>>> if (err)
> >>>> goto out_err;
> >>>> err = ceph_pagelist_encode_string(pagelist,
> >>> This is wrong. In both cases it reserves pages for two string-like
> >>> things, str_size + 4 each, for a total of str_size1 + str_size2 + 8.
> >> Ah, yes. Could we change to sizeof(u32) + sizeof(u32) instead?
> > No, that would be unnecessary churn.
> Probably we can change the first reservation size from PAGE_SIZE to
> sizeof(u32)
> in ceph_pre_init_acls() because actually we encode int type 2 or 1 using
> ceph_pagelist_encode_32(). Am I missing something?
It wouldn't change anything. It is a fresly created pagelist (i.e.
a list of pages). The granularity is a page, so even if you reserve
just one byte, it would still allocate a page.
Thanks,
Ilya
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
