Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752159AbcDLE0d (ORCPT ); Tue, 12 Apr 2016 00:26:33 -0400 Received: from mga14.intel.com ([192.55.52.115]:2126 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751247AbcDLE0c (ORCPT ); Tue, 12 Apr 2016 00:26:32 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.24,472,1455004800"; d="scan'208";a="943209384" Date: Tue, 12 Apr 2016 07:26:27 +0300 From: Jarkko Sakkinen To: Jason Gunthorpe Cc: Peter Huewe , Marcel Selhorst , "moderated list:TPM DEVICE DRIVER" , open list Subject: Re: [PATCH] tpm: fix crash in tpm_tis deinitialization Message-ID: <20160412042627.GA6608@intel.com> References: <1460390720-9509-1-git-send-email-jarkko.sakkinen@linux.intel.com> <20160411174037.GA371@obsidianresearch.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20160411174037.GA371@obsidianresearch.com> Organization: Intel Finland Oy - BIC 0357606-4 - Westendinkatu 7, 02160 Espoo User-Agent: Mutt/1.5.24 (2015-08-30) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2288 Lines: 59 On Mon, Apr 11, 2016 at 11:40:37AM -0600, Jason Gunthorpe wrote: > On Mon, Apr 11, 2016 at 07:05:20PM +0300, Jarkko Sakkinen wrote: > > rmmod crashes the driver because tpm_chip_unregister() already sets ops > > to NULL. This commit fixes the issue by moving tpm2_shutdown() to > > tpm_chip_unregister(). This commit is also cleanup because it removes > > duplicate code from tpm_crb and tpm_tis to the core. > > > > Signed-off-by: Jarkko Sakkinen > > Fixes: 4d3eac5e156a ("tpm: Provide strong locking for device removal") > > drivers/char/tpm/tpm-chip.c | 3 +++ > > drivers/char/tpm/tpm_crb.c | 3 --- > > drivers/char/tpm/tpm_tis.c | 3 --- > > 3 files changed, 3 insertions(+), 6 deletions(-) > > > > diff --git a/drivers/char/tpm/tpm-chip.c b/drivers/char/tpm/tpm-chip.c > > index f62c851..2642cca 100644 > > +++ b/drivers/char/tpm/tpm-chip.c > > @@ -361,6 +361,9 @@ void tpm_chip_unregister(struct tpm_chip *chip) > > if (!(chip->flags & TPM_CHIP_FLAG_REGISTERED)) > > return; > > > > + if (chip->flags & TPM_CHIP_FLAG_TPM2) > > + tpm2_shutdown(chip, TPM2_SU_CLEAR); > > + > > if (!(chip->flags & TPM_CHIP_FLAG_TPM2)) > > sysfs_remove_link(&chip->dev.parent->kobj, "ppi"); > > This needs to be after ops is fenced, something like this. I would appreciate a supporting argument. I guess the argument here is that this will prevent user space from issuing TPM commands after the shutdown command has been sent? /Jarkko > diff --git a/drivers/char/tpm/tpm-chip.c b/drivers/char/tpm/tpm-chip.c > index 2642cca05cac..2ea2f1561e59 100644 > --- a/drivers/char/tpm/tpm-chip.c > +++ b/drivers/char/tpm/tpm-chip.c > @@ -269,6 +269,8 @@ static void tpm_del_char_device(struct tpm_chip *chip) > > /* Make the driver uncallable. */ > down_write(&chip->ops_sem); > + if (chip->flags & TPM_CHIP_FLAG_TPM2) > + tpm2_shutdown(chip, TPM2_SU_CLEAR); > chip->ops = NULL; > up_write(&chip->ops_sem); > } > @@ -361,9 +363,6 @@ void tpm_chip_unregister(struct tpm_chip *chip) > if (!(chip->flags & TPM_CHIP_FLAG_REGISTERED)) > return; > > - if (chip->flags & TPM_CHIP_FLAG_TPM2) > - tpm2_shutdown(chip, TPM2_SU_CLEAR); > - > if (!(chip->flags & TPM_CHIP_FLAG_TPM2)) > sysfs_remove_link(&chip->dev.parent->kobj, "ppi"); >