Re: [PATCH v11 00/14] TPM IRQ fixes

[Jarkko Sakkinen <jarkko@kernel.org>]

On Fri, Apr 21, 2023 at 10:34:21PM +0300, Jarkko Sakkinen wrote:
> On Fri, Apr 21, 2023 at 07:50:14PM +0300, Jarkko Sakkinen wrote:
> > On Sun, Mar 19, 2023 at 03:53:38PM +0200, Jarkko Sakkinen wrote:
> > > On Thu, Mar 16, 2023 at 06:18:11PM +0100, Michael Niewöhner wrote:
> > > > On Fri, 2023-02-24 at 15:48 +0100, Lino Sanfilippo wrote:
> > > > > On 24.11.22 at 14:55, Lino Sanfilippo wrote:
> > > > > > From: Lino Sanfilippo <l.sanfilippo@kunbus.com>
> > > > > > 
> > > > > > This series enables IRQ support for the TPM TIS core. For this reason a
> > > > > > number of bugfixes around the interrupt handling are required (patches 1 to
> > > > > > 5).
> > > > > > 
> > > > > > Patch 6 takes into account that according to the TPM Interface
> > > > > > Specification stsValid and commandRead interrupts might not be supported
> > > > > > by the hardware. For this reason the supported interrupts are first queried
> > > > > > and stored. Then wait_for_tpm_stat() is adjusted to not wait for status
> > > > > > changes that are not reported by interrupts.
> > > > > > 
> > > > > > Patch 7 moves the interrupt flag checks into an own function as suggested
> > > > > > by Jarkko.
> > > > > > 
> > > > > > Patch 8 Removes the possibility that tpm_tis_data->locality can be changed
> > > > > > at driver runtime so this variable can be read without the need to protect
> > > > > > it against concurrent modification.
> > > > > > 
> > > > > > Patch 9 addresses the issue with concurrent locality handling:
> > > > > > Since the interrupt handler writes the interrupt status registers it needs
> > > > > > to hold the locality. However it runs concurrently to the thread which
> > > > > > triggered the interrupt (e.g. by reading or writing data to the TPM). So
> > > > > > it must take care when claiming and releasing the locality itself,
> > > > > > because it may race with the concurrent running thread which also claims
> > > > > > and releases the locality.
> > > > > > To avoid that both interrupt and concurrent running thread interfere with
> > > > > > each other a locality counter is used which guarantees that at any time
> > > > > > the locality is held as long as it is required by one of both execution
> > > > > > paths.
> > > > > > 
> > > > > > Patch 10 implements the request of a threaded interrupt handler. This is
> > > > > > needed since SPI uses a mutex for data transmission and since we access the
> > > > > > interrupt status register via SPI in the irq handler we need a sleepable
> > > > > > context.
> > > > > > 
> > > > > > Patch 11 makes sure that writes to the interrupt register are effective if
> > > > > > done in the interrupt handler.
> > > > > > 
> > > > > > Patch 12 makes sure that writes to the interrupt and INTERRUPT_VECTOR
> > > > > > and INTERRUPT_ENABLE registers are effective by holding the locality.
> > > > > > 
> > > > > > Patch 13 makes sure that the TPM is properly initialized after power cycle
> > > > > > before the irq test is done.
> > > > > > 
> > > > > > Patch 14 enables the test for interrupts by setting the required flag
> > > > > > before the test is executed.
> > > > > > 
> > > > > > Changes in v11:
> > > > > > - adjust patches 4,5 and 14 slightly to void the consecutive removal and
> > > > > >   re-addition of the "ret" variable in tpm_tis_gen_interrupt()
> > > > > > 
> > > > > > Changes in v10:
> > > > > > - fix typo in subject line as pointed out by Jason Andryuk
> > > > > > - improve patch "tpm, tpm_tis: Claim locality before writing interrupt
> > > > > >   registers" by extending the scope with held locality". For this reason
> > > > > >   the "Reviewed-by" tag by Jarko and the "Tested-by" tag by Michael have
> > > > > >   been removed.
> > > > > > - add fix to avoid TPM_RC_INITIALIZE after power cycle when testing irqs
> > > > > >   (PATCH 13)
> > > > > > - add fix to restore the old interrupt vector at error (PATCH 4)
> > > > > > 
> > > > > > 
> > > > > > Changes in v9:
> > > > > > - add a fix for an issue when interrupts are reenabled on resume (PATCH 11)
> > > > > > - improve the commit message for patch 8 as requested by Jarkko
> > > > > > - improved functions naming
> > > > > > - changed patch 12 (tpm, tpm_tis: Enable interrupt test) to not delete the
> > > > > >   TPM_CHIP_FLAG_IRQ flag any more when tpm2_get_tpm_pt() fails. Due to this
> > > > > >   change the 'Tested-by' tag from Michael and the 'Reviewed-by:' tag from
> > > > > >   Jarko has been removed
> > > > > > 
> > > > > > Changes in v8:
> > > > > > - tpm_tis_data->locality is not changed at runtime any more so that it can
> > > > > > be read without any protection against concurrent modification.
> > > > > > - add missing brackets as pointed out by Jason Andryuk
> > > > > > 
> > > > > > Changes in v7:
> > > > > > - moved interrupt flag checks into an own function as suggested by Jarkko
> > > > > > - added "Tested-by" tags for Tests from Michael Niewöhner
> > > > > > - fixed one comment
> > > > > > 
> > > > > > Changes in v6:
> > > > > > - set TPM_TIS_IRQ_TESTED in flag member of the tpm_tis_data struct instead
> > > > > > in an own bitfield
> > > > > > - improve commit messages
> > > > > > - use int_mask instead of irqs_in_use as variable name
> > > > > > - use sts_mask instead of active_irqs as variable name
> > > > > > - squash patch 5 and 6
> > > > > > - prefix functions with tpm_tis_
> > > > > > - remove "fixes" tag
> > > > > > 
> > > > > > Changes in v5:
> > > > > > - improve commit message of patch 1 as requested by Jarko
> > > > > > - drop patch that makes locality handling simpler by only claiming it at
> > > > > >   driver startup and releasing it at driver shutdown (requested by Jarko)
> > > > > > - drop patch that moves the interrupt test from tpm_tis_send()
> > > > > >   to tmp_tis_probe_irq_single() as requested by Jarko
> > > > > > - add patch to make locality handling threadsafe so that it can also be
> > > > > >   done by the irq handler
> > > > > > - separate logical changes into own patches
> > > > > > - always request threaded interrupt handler
> > > > > > 
> > > > > > Changes in v4:
> > > > > > - only request threaded irq in case of SPI as requested by Jarko.
> > > > > > - reimplement patch 2 to limit locality handling changes to the TIS core.
> > > > > > - separate fixes from cleanups as requested by Jarko.
> > > > > > - rephrase commit messages
> > > > > > 
> > > > > > Changes in v3:
> > > > > > - fixed compiler error reported by kernel test robot
> > > > > > - rephrased commit message as suggested by Jarko Sakkinen
> > > > > > - added Reviewed-by tag
> > > > > > 
> > > > > > Changes in v2:
> > > > > > - rebase against 5.12
> > > > > > - free irq on error path
> > > > > > 
> > > > > > 
> > > > > > Lino Sanfilippo (14):
> > > > > >   tpm, tpm_tis: Avoid cache incoherency in test for interrupts
> > > > > >   tpm, tpm_tis: Claim locality before writing TPM_INT_ENABLE register
> > > > > >   tpm, tpm_tis: Disable interrupts if tpm_tis_probe_irq() failed
> > > > > >   tpm, tpm_tis: Do not skip reset of original interrupt vector
> > > > > >   tpm, tpm_tis: Claim locality before writing interrupt registers
> > > > > >   tpm, tpm_tis: Only handle supported interrupts
> > > > > >   tpm, tpm_tis: Move interrupt mask checks into own function
> > > > > >   tpm, tpm_tis: do not check for the active locality in interrupt
> > > > > >     handler
> > > > > >   tpm, tpm: Implement usage counter for locality
> > > > > >   tpm, tpm_tis: Request threaded interrupt handler
> > > > > >   tpm, tpm_tis: Claim locality in interrupt handler
> > > > > >   tpm, tpm_tis: Claim locality when interrupts are reenabled on resume
> > > > > >   tpm, tpm_tis: startup chip before testing for interrupts
> > > > > >   tpm, tpm_tis: Enable interrupt test
> > > > > > 
> > > > > >  drivers/char/tpm/tpm-chip.c     |  38 ++--
> > > > > >  drivers/char/tpm/tpm.h          |   1 +
> > > > > >  drivers/char/tpm/tpm_tis.c      |   2 +-
> > > > > >  drivers/char/tpm/tpm_tis_core.c | 299 ++++++++++++++++++++------------
> > > > > >  drivers/char/tpm/tpm_tis_core.h |   5 +-
> > > > > >  5 files changed, 214 insertions(+), 131 deletions(-)
> > > > > > 
> > > > > 
> > > > > Hi Jarkko,
> > > > > 
> > > > > its been a while now since the review of this series has been completed. Will
> > > > > it be merged in the
> > > > > near future? Or is there anything left to do (from my side)?
> > > > > 
> > > > > Regards,
> > > > > Lino
> > > > > 
> > > > 
> > > > @Jarkko ping ;)
> > > > 
> > > 
> > > Thanks for reminding. I was wondering this week what was the situation
> > > but latest version I had bookmarked from lore was 10.
> > > 
> > > Please ping earlier! I'll get on testing this, apologies.
> > 
> > I tested this with libvirt/QEMU/swtpm and did the following tests:
> > 
> > 1. TPM 1.2 suspend/resume.
> > 2. TPM 2.0 kselftest.
> > 3. TPM 2.0 suspend/resume + kselftest.
> > 
> > I see no issues so I can pick this for my pull request.
> > 
> > Tests were performed on top of v6.3-rc7.
> > 
> > For all:
> > 
> > Tested-by: Jarkko Sakkinen <jarkko@kernel.org>
> 
> Changes are in my tree and should be soon mirrored to linux-next.
> 
> Please check for any possible abnormalities, I tried to be careful.

I'd also expect that there will appear some yet to be know issues
(if not, even better ofc) but there is full cycle to fix them.

BR, Jarkko