Requesting backport for fc20c523211 (cifs: fixes for get_inode_info)

Requesting backport for fc20c523211 (cifs: fixes for get_inode_info)

[Meetakshi Setiya]

commit fc20c523211a38b87fc850a959cb2149e4fd64b0 upstream
cifs: fixes for get_inode_info
requesting backport to 6.8.x, 6.6.x, 6.5.x and 6.1.x

This patch fixes memory leaks, adds error checking, and performs some important
code modifications to the changes introduced by patch 2 of this patch series:
https://lore.kernel.org/stable/CAFTVevX6=4qFo6nwV14sCnfPRO9yb9q+YsP3XPaHMsP08E05iQ@mail.gmail.com/
commit ffceb7640cbfe6ea60e7769e107451d63a2fe3d3
(smb: client: do not defer close open handles to deleted files)

This patch and the three patches in the mails that precede this are related and
fix an important customer reported bug on the linux smb client (explained in the
mail for patch 1). Patches 2, 3 and 4 are meant to fix whatever regressions were
introduced/exposed by patch 1.
The patches have to be applied in the mentioned order and should be backported
together.

Patch 1: https://lore.kernel.org/stable/CAFTVevWEnEDAQbw59N-R03ppFgqa3qwTySfn61-+T4Vodq97Gw@mail.gmail.com/
Patch 2: https://lore.kernel.org/stable/CAFTVevX6=4qFo6nwV14sCnfPRO9yb9q+YsP3XPaHMsP08E05iQ@mail.gmail.com/
Patch 3: https://lore.kernel.org/stable/CAFTVevX6Yzjm40EoGZzex6G-f3T-YNG2CZMAuy=fBSwx9hm8Jw@mail.gmail.com/
Patch 4: The current patch

Thanks
Meetakshi

Re: Requesting backport for fc20c523211 (cifs: fixes for get_inode_info)

[Greg KH]

On Wed, Apr 03, 2024 at 12:34:43PM +0530, Meetakshi Setiya wrote:
> commit fc20c523211a38b87fc850a959cb2149e4fd64b0 upstream
> cifs: fixes for get_inode_info
> requesting backport to 6.8.x, 6.6.x, 6.5.x and 6.1.x
> 
> This patch fixes memory leaks, adds error checking, and performs some important
> code modifications to the changes introduced by patch 2 of this patch series:
> https://lore.kernel.org/stable/CAFTVevX6=4qFo6nwV14sCnfPRO9yb9q+YsP3XPaHMsP08E05iQ@mail.gmail.com/
> commit ffceb7640cbfe6ea60e7769e107451d63a2fe3d3
> (smb: client: do not defer close open handles to deleted files)
> 
> This patch and the three patches in the mails that precede this are related and
> fix an important customer reported bug on the linux smb client (explained in the
> mail for patch 1). Patches 2, 3 and 4 are meant to fix whatever regressions were
> introduced/exposed by patch 1.
> The patches have to be applied in the mentioned order and should be backported
> together.

Then PLEASE send this as a patch series, as picking patches out of
emails that arrive in random order in a "correct" way is tough, if not
impossible for us to do.

Please send these as a backported set of patches, OR as a list of
"cherry-pick these git ids in this order" type of thing.  But spreading
it out over 4 emails just does not work, and is very very confusing.

Think about it, would you want to recieve these 4 emails and have to try
to guess which one is applied before which?

Please fix this up and resend it in an easier way.

thanks,

greg k-h

Re: Requesting backport for fc20c523211 (cifs: fixes for get_inode_info)

[Steve French]

On Fri, Apr 5, 2024 at 1:35 AM Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Wed, Apr 03, 2024 at 12:34:43PM +0530, Meetakshi Setiya wrote:
> > commit fc20c523211a38b87fc850a959cb2149e4fd64b0 upstream
> > cifs: fixes for get_inode_info
> > requesting backport to 6.8.x, 6.6.x, 6.5.x and 6.1.x
> >
> > This patch fixes memory leaks, adds error checking, and performs some important
> > code modifications to the changes introduced by patch 2 of this patch series:
> > https://lore.kernel.org/stable/CAFTVevX6=4qFo6nwV14sCnfPRO9yb9q+YsP3XPaHMsP08E05iQ@mail.gmail.com/
> > commit ffceb7640cbfe6ea60e7769e107451d63a2fe3d3
> > (smb: client: do not defer close open handles to deleted files)
> >
> > This patch and the three patches in the mails that precede this are related and
> > fix an important customer reported bug on the linux smb client (explained in the
> > mail for patch 1). Patches 2, 3 and 4 are meant to fix whatever regressions were
> > introduced/exposed by patch 1.
> > The patches have to be applied in the mentioned order and should be backported
> > together.
>
> Then PLEASE send this as a patch series, as picking patches out of
> emails that arrive in random order in a "correct" way is tough, if not
> impossible for us to do.
>
> Please send these as a backported set of patches, OR as a list of
> "cherry-pick these git ids in this order" type of thing.  But spreading
> it out over 4 emails just does not work, and is very very confusing.

To make it easier, I recommend we wait a few days on this as there is
one more important fix for this series that was recently found (by Paulo)
and I haven't sent to Linus yet - then can send the complete set
for at least 6.8 and 6.6 stable.  Do you prefer a separate email
for the 6.8 version of these, and another for the 6.6 rebased
version of the series - or all as one email? AFAIK she hasn't
rebased for 6.1LTS.



-- 
Thanks,

Steve

Re: Requesting backport for fc20c523211 (cifs: fixes for get_inode_info)

[Greg KH]

On Wed, Apr 10, 2024 at 12:51:32AM -0500, Steve French wrote:
> On Fri, Apr 5, 2024 at 1:35 AM Greg KH <gregkh@linuxfoundation.org> wrote:
> >
> > On Wed, Apr 03, 2024 at 12:34:43PM +0530, Meetakshi Setiya wrote:
> > > commit fc20c523211a38b87fc850a959cb2149e4fd64b0 upstream
> > > cifs: fixes for get_inode_info
> > > requesting backport to 6.8.x, 6.6.x, 6.5.x and 6.1.x
> > >
> > > This patch fixes memory leaks, adds error checking, and performs some important
> > > code modifications to the changes introduced by patch 2 of this patch series:
> > > https://lore.kernel.org/stable/CAFTVevX6=4qFo6nwV14sCnfPRO9yb9q+YsP3XPaHMsP08E05iQ@mail.gmail.com/
> > > commit ffceb7640cbfe6ea60e7769e107451d63a2fe3d3
> > > (smb: client: do not defer close open handles to deleted files)
> > >
> > > This patch and the three patches in the mails that precede this are related and
> > > fix an important customer reported bug on the linux smb client (explained in the
> > > mail for patch 1). Patches 2, 3 and 4 are meant to fix whatever regressions were
> > > introduced/exposed by patch 1.
> > > The patches have to be applied in the mentioned order and should be backported
> > > together.
> >
> > Then PLEASE send this as a patch series, as picking patches out of
> > emails that arrive in random order in a "correct" way is tough, if not
> > impossible for us to do.
> >
> > Please send these as a backported set of patches, OR as a list of
> > "cherry-pick these git ids in this order" type of thing.  But spreading
> > it out over 4 emails just does not work, and is very very confusing.
> 
> To make it easier, I recommend we wait a few days on this as there is
> one more important fix for this series that was recently found (by Paulo)
> and I haven't sent to Linus yet - then can send the complete set
> for at least 6.8 and 6.6 stable.  Do you prefer a separate email
> for the 6.8 version of these, and another for the 6.6 rebased
> version of the series - or all as one email? AFAIK she hasn't
> rebased for 6.1LTS.

If the versions are different, yes, individual series are appreciated.
If they are identical, one is fine.

thanks,

greg k-h