Обсуждение: Remove unnecessary code rom be_lo_put()
Hi, I noticed that a permission check is performed in be_lo_put() just after returning inv_open(), but teh permission should be already checked in inv_open(), so I think we can remove this part of codes. I attached a patch for this fix. Regards, Yugo Nagata -- Yugo NAGATA <nagata@sraoss.co.jp>
Вложения
On 24.04.24 11:59, Yugo NAGATA wrote: > I noticed that a permission check is performed in be_lo_put() > just after returning inv_open(), but teh permission should be > already checked in inv_open(), so I think we can remove this > part of codes. I attached a patch for this fix. Yes, I think you are right. This check was added in 8d9881911f0, but then the refactoring in ae20b23a9e7 should probably have removed it.
Peter Eisentraut <peter@eisentraut.org> writes: > On 24.04.24 11:59, Yugo NAGATA wrote: >> I noticed that a permission check is performed in be_lo_put() >> just after returning inv_open(), but teh permission should be >> already checked in inv_open(), so I think we can remove this >> part of codes. I attached a patch for this fix. > Yes, I think you are right. I agree. Do you want to do the honors? regards, tom lane
On Wed, Apr 24, 2024 at 09:25:09AM -0400, Tom Lane wrote: > I agree. Do you want to do the honors? Good catch. The same check happens when the object is opened. Note that you should be able to remove utils/acl.h at the top of be-fsstubs.c as this would remove the last piece of code that does an ACL check in this file. No objections with doing that now, removing this code. -- Michael
Вложения
On 25.04.24 01:50, Michael Paquier wrote: > On Wed, Apr 24, 2024 at 09:25:09AM -0400, Tom Lane wrote: >> I agree. Do you want to do the honors? > > Good catch. The same check happens when the object is opened. Note > that you should be able to remove utils/acl.h at the top of > be-fsstubs.c as this would remove the last piece of code that does an > ACL check in this file. No objections with doing that now, removing > this code. utils/acl.h is still needed for object_ownercheck() called in be_lo_unlink().
On 24.04.24 15:25, Tom Lane wrote: > Peter Eisentraut <peter@eisentraut.org> writes: >> On 24.04.24 11:59, Yugo NAGATA wrote: >>> I noticed that a permission check is performed in be_lo_put() >>> just after returning inv_open(), but teh permission should be >>> already checked in inv_open(), so I think we can remove this >>> part of codes. I attached a patch for this fix. > >> Yes, I think you are right. > > I agree. Do you want to do the honors? done
On Thu, 25 Apr 2024 10:26:41 +0200 Peter Eisentraut <peter@eisentraut.org> wrote: > On 24.04.24 15:25, Tom Lane wrote: > > Peter Eisentraut <peter@eisentraut.org> writes: > >> On 24.04.24 11:59, Yugo NAGATA wrote: > >>> I noticed that a permission check is performed in be_lo_put() > >>> just after returning inv_open(), but teh permission should be > >>> already checked in inv_open(), so I think we can remove this > >>> part of codes. I attached a patch for this fix. > > > >> Yes, I think you are right. > > > > I agree. Do you want to do the honors? > > done > Thank you! Regards, Yugo Nagata -- Yugo NAGATA <nagata@sraoss.co.jp>