[GIT,PULL] tee subsys fixes for v5.4 (take two)
mbox series

Message ID 20191115105353.GA26176@jax
State New
Headers show
Series
  • [GIT,PULL] tee subsys fixes for v5.4 (take two)
Related show

Pull-request

git://git.linaro.org/people/jens.wiklander/linux-tee.git/ tags/tee-fixes-for-v5.4

Message

Jens Wiklander Nov. 15, 2019, 10:53 a.m. UTC
Hello arm-soc maintainers,

Please pull these OP-TEE driver fixes. There's one user-after-free issue if
in the error handling path when the OP-TEE driver is initializing. There's
also one fix to to register dynamically allocated shared memory needed by
kernel clients communicating with secure world via memory references.

"tee: optee: Fix dynamic shm pool allocations" is now from version 2 which
includes a fix up with a small but vital dependency.

If you think it's too late for v5.4 please queue this for v5.5 instead.

Thanks,
Jens


The following changes since commit 4f5cafb5cb8471e54afdc9054d973535614f7675:

  Linux 5.4-rc3 (2019-10-13 16:37:36 -0700)

are available in the Git repository at:

  git://git.linaro.org/people/jens.wiklander/linux-tee.git/ tags/tee-fixes-for-v5.4

for you to fetch changes up to 03212e347f9443e524d6383c6806ac08295c1fb0:

  tee: optee: fix device enumeration error handling (2019-11-15 11:31:24 +0100)

----------------------------------------------------------------
Two OP-TE driver fixes:
- Add proper cleanup on optee_enumerate_devices() failure
- Make sure to register kernel allocations of dynamic shared memory

----------------------------------------------------------------
Jens Wiklander (1):
      tee: optee: fix device enumeration error handling

Sumit Garg (1):
      tee: optee: Fix dynamic shm pool allocations

 drivers/tee/optee/call.c     |  7 +++++++
 drivers/tee/optee/core.c     | 20 ++++++++++++--------
 drivers/tee/optee/shm_pool.c | 12 +++++++++++-
 3 files changed, 30 insertions(+), 9 deletions(-)

Comments

Olof Johansson Nov. 16, 2019, 11:40 p.m. UTC | #1
On Fri, Nov 15, 2019 at 11:53:53AM +0100, Jens Wiklander wrote:
> Hello arm-soc maintainers,
> 
> Please pull these OP-TEE driver fixes. There's one user-after-free issue if
> in the error handling path when the OP-TEE driver is initializing. There's
> also one fix to to register dynamically allocated shared memory needed by
> kernel clients communicating with secure world via memory references.
> 
> "tee: optee: Fix dynamic shm pool allocations" is now from version 2 which
> includes a fix up with a small but vital dependency.
> 
> If you think it's too late for v5.4 please queue this for v5.5 instead.

Hi,

I noticed you based this on -rc3 -- all our other branches are on -rc2 or
older.

Anyway, I brought this in to the fixes branch, it's the only thing we have
queued up at this time so I'll give it a few days in -next before I send it in.


-Olof
Jens Wiklander Nov. 17, 2019, 2:22 p.m. UTC | #2
On Sun, Nov 17, 2019 at 12:45 AM Olof Johansson <olof@lixom.net> wrote:
>
> On Fri, Nov 15, 2019 at 11:53:53AM +0100, Jens Wiklander wrote:
> > Hello arm-soc maintainers,
> >
> > Please pull these OP-TEE driver fixes. There's one user-after-free issue if
> > in the error handling path when the OP-TEE driver is initializing. There's
> > also one fix to to register dynamically allocated shared memory needed by
> > kernel clients communicating with secure world via memory references.
> >
> > "tee: optee: Fix dynamic shm pool allocations" is now from version 2 which
> > includes a fix up with a small but vital dependency.
> >
> > If you think it's too late for v5.4 please queue this for v5.5 instead.
>
> Hi,
>
> I noticed you based this on -rc3 -- all our other branches are on -rc2 or
> older.

I'm sorry, I thought -rc3 was old enough. I'll stick to -rc2 or older
in next time.

>
> Anyway, I brought this in to the fixes branch, it's the only thing we have
> queued up at this time so I'll give it a few days in -next before I send it in.

Great.

Thanks,
Jens

>
>
> -Olof
Jens Wiklander Dec. 6, 2019, 9:57 a.m. UTC | #3
Hi Olof,

On Sun, Nov 17, 2019 at 3:22 PM Jens Wiklander
<jens.wiklander@linaro.org> wrote:
>
> On Sun, Nov 17, 2019 at 12:45 AM Olof Johansson <olof@lixom.net> wrote:
> >
> > On Fri, Nov 15, 2019 at 11:53:53AM +0100, Jens Wiklander wrote:
> > > Hello arm-soc maintainers,
> > >
> > > Please pull these OP-TEE driver fixes. There's one user-after-free issue if
> > > in the error handling path when the OP-TEE driver is initializing. There's
> > > also one fix to to register dynamically allocated shared memory needed by
> > > kernel clients communicating with secure world via memory references.
> > >
> > > "tee: optee: Fix dynamic shm pool allocations" is now from version 2 which
> > > includes a fix up with a small but vital dependency.
> > >
> > > If you think it's too late for v5.4 please queue this for v5.5 instead.
> >
> > Hi,
> >
> > I noticed you based this on -rc3 -- all our other branches are on -rc2 or
> > older.
>
> I'm sorry, I thought -rc3 was old enough. I'll stick to -rc2 or older
> in next time.
>
> >
> > Anyway, I brought this in to the fixes branch, it's the only thing we have
> > queued up at this time so I'll give it a few days in -next before I send it in.

It looks like the two patches in this pull request
(https://git.linaro.org/people/jens.wiklander/linux-tee.git/tag/?h=tee-fixes-for-v5.4)
are still in -next and haven't got any further. Is there anything
wrong? Something I should fix?

Thanks,
Jens
Olof Johansson Dec. 6, 2019, 4:29 p.m. UTC | #4
On Fri, Dec 06, 2019 at 10:57:47AM +0100, Jens Wiklander wrote:
> Hi Olof,
> 
> On Sun, Nov 17, 2019 at 3:22 PM Jens Wiklander
> <jens.wiklander@linaro.org> wrote:
> >
> > On Sun, Nov 17, 2019 at 12:45 AM Olof Johansson <olof@lixom.net> wrote:
> > >
> > > On Fri, Nov 15, 2019 at 11:53:53AM +0100, Jens Wiklander wrote:
> > > > Hello arm-soc maintainers,
> > > >
> > > > Please pull these OP-TEE driver fixes. There's one user-after-free issue if
> > > > in the error handling path when the OP-TEE driver is initializing. There's
> > > > also one fix to to register dynamically allocated shared memory needed by
> > > > kernel clients communicating with secure world via memory references.
> > > >
> > > > "tee: optee: Fix dynamic shm pool allocations" is now from version 2 which
> > > > includes a fix up with a small but vital dependency.
> > > >
> > > > If you think it's too late for v5.4 please queue this for v5.5 instead.
> > >
> > > Hi,
> > >
> > > I noticed you based this on -rc3 -- all our other branches are on -rc2 or
> > > older.
> >
> > I'm sorry, I thought -rc3 was old enough. I'll stick to -rc2 or older
> > in next time.
> >
> > >
> > > Anyway, I brought this in to the fixes branch, it's the only thing we have
> > > queued up at this time so I'll give it a few days in -next before I send it in.
> 
> It looks like the two patches in this pull request
> (https://git.linaro.org/people/jens.wiklander/linux-tee.git/tag/?h=tee-fixes-for-v5.4)
> are still in -next and haven't got any further. Is there anything
> wrong? Something I should fix?

They were in our fixes branch and didn't go in yet, but will shortly.


-Olof