linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2] tpm: Return the actual size when receiving an unsupported command
@ 2018-06-20  9:55 Ricardo Schwarzmeier
  2018-06-20 14:14 ` kbuild test robot
  2018-06-21 16:48 ` Jarkko Sakkinen
  0 siblings, 2 replies; 3+ messages in thread
From: Ricardo Schwarzmeier @ 2018-06-20  9:55 UTC (permalink / raw)
  To: linux-integrity
  Cc: linux-kernel, peterhuewe, jarkko.sakkinen, jgg, arnd, gregkh,
	Ricardo Schwarzmeier

The command size is computed for the header anyway.
It can as well indicate the correct number of bytes to read from the driver.
Fixes: 095531f891e6 ("tpm: return a TPM_RC_COMMAND_CODE response if command is not implemented")

Signed-off-by: Ricardo Schwarzmeier <Ricardo.Schwarzmeier@infineon.com>
---
 drivers/char/tpm/tpm-interface.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/char/tpm/tpm-interface.c b/drivers/char/tpm/tpm-interface.c
index e32f6e85dc6d..0923e81505a5 100644
--- a/drivers/char/tpm/tpm-interface.c
+++ b/drivers/char/tpm/tpm-interface.c
@@ -423,7 +423,7 @@ static ssize_t tpm_try_transmit(struct tpm_chip *chip,
 		header->tag = cpu_to_be16(TPM2_ST_NO_SESSIONS);
 		header->return_code = cpu_to_be32(TPM2_RC_COMMAND_CODE |
 						  TSS2_RESMGR_TPM_RC_LAYER);
-		return bufsiz;
+		return header->length;
 	}
 
 	if (bufsiz > TPM_BUFSIZE)
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH v2] tpm: Return the actual size when receiving an unsupported command
  2018-06-20  9:55 [PATCH v2] tpm: Return the actual size when receiving an unsupported command Ricardo Schwarzmeier
@ 2018-06-20 14:14 ` kbuild test robot
  2018-06-21 16:48 ` Jarkko Sakkinen
  1 sibling, 0 replies; 3+ messages in thread
From: kbuild test robot @ 2018-06-20 14:14 UTC (permalink / raw)
  To: Ricardo Schwarzmeier
  Cc: kbuild-all, linux-integrity, linux-kernel, peterhuewe,
	jarkko.sakkinen, jgg, arnd, gregkh, Ricardo Schwarzmeier

Hi Ricardo,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on char-misc/char-misc-testing]
[also build test WARNING on v4.18-rc1 next-20180620]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Ricardo-Schwarzmeier/tpm-Return-the-actual-size-when-receiving-an-unsupported-command/20180620-210928
reproduce:
        # apt-get install sparse
        make ARCH=x86_64 allmodconfig
        make C=1 CF=-D__CHECK_ENDIAN__


sparse warnings: (new ones prefixed by >>)

>> drivers/char/tpm/tpm-interface.c:426:30: sparse: incorrect type in return expression (different base types) @@    expected long @@    got restricted __be32 [usertylong @@
   drivers/char/tpm/tpm-interface.c:426:30:    expected long
   drivers/char/tpm/tpm-interface.c:426:30:    got restricted __be32 [usertype] length
   drivers/char/tpm/tpm-interface.c:565:34: sparse: expression using sizeof(void)
   drivers/char/tpm/tpm-interface.c:565:34: sparse: expression using sizeof(void)
   drivers/char/tpm/tpm-interface.c:1258:31: sparse: expression using sizeof(void)

vim +426 drivers/char/tpm/tpm-interface.c

   401	
   402	static ssize_t tpm_try_transmit(struct tpm_chip *chip,
   403					struct tpm_space *space,
   404					u8 *buf, size_t bufsiz,
   405					unsigned int flags)
   406	{
   407		struct tpm_output_header *header = (void *)buf;
   408		int rc;
   409		ssize_t len = 0;
   410		u32 count, ordinal;
   411		unsigned long stop;
   412		bool need_locality;
   413	
   414		rc = tpm_validate_command(chip, space, buf, bufsiz);
   415		if (rc == -EINVAL)
   416			return rc;
   417		/*
   418		 * If the command is not implemented by the TPM, synthesize a
   419		 * response with a TPM2_RC_COMMAND_CODE return for user-space.
   420		 */
   421		if (rc == -EOPNOTSUPP) {
   422			header->length = cpu_to_be32(sizeof(*header));
   423			header->tag = cpu_to_be16(TPM2_ST_NO_SESSIONS);
   424			header->return_code = cpu_to_be32(TPM2_RC_COMMAND_CODE |
   425							  TSS2_RESMGR_TPM_RC_LAYER);
 > 426			return header->length;
   427		}
   428	
   429		if (bufsiz > TPM_BUFSIZE)
   430			bufsiz = TPM_BUFSIZE;
   431	
   432		count = be32_to_cpu(*((__be32 *) (buf + 2)));
   433		ordinal = be32_to_cpu(*((__be32 *) (buf + 6)));
   434		if (count == 0)
   435			return -ENODATA;
   436		if (count > bufsiz) {
   437			dev_err(&chip->dev,
   438				"invalid count value %x %zx\n", count, bufsiz);
   439			return -E2BIG;
   440		}
   441	
   442		if (!(flags & TPM_TRANSMIT_UNLOCKED))
   443			mutex_lock(&chip->tpm_mutex);
   444	
   445	
   446		if (chip->ops->clk_enable != NULL)
   447			chip->ops->clk_enable(chip, true);
   448	
   449		/* Store the decision as chip->locality will be changed. */
   450		need_locality = chip->locality == -1;
   451	
   452		if (!(flags & TPM_TRANSMIT_RAW) && need_locality) {
   453			rc = tpm_request_locality(chip);
   454			if (rc < 0)
   455				goto out_no_locality;
   456		}
   457	
   458		if (chip->dev.parent)
   459			pm_runtime_get_sync(chip->dev.parent);
   460	
   461		rc = tpm2_prepare_space(chip, space, ordinal, buf);
   462		if (rc)
   463			goto out;
   464	
   465		rc = chip->ops->send(chip, buf, count);
   466		if (rc < 0) {
   467			if (rc != -EPIPE)
   468				dev_err(&chip->dev,
   469					"%s: tpm_send: error %d\n", __func__, rc);
   470			goto out;
   471		}
   472	
   473		if (chip->flags & TPM_CHIP_FLAG_IRQ)
   474			goto out_recv;
   475	
   476		if (chip->flags & TPM_CHIP_FLAG_TPM2)
   477			stop = jiffies + tpm2_calc_ordinal_duration(chip, ordinal);
   478		else
   479			stop = jiffies + tpm_calc_ordinal_duration(chip, ordinal);
   480		do {
   481			u8 status = chip->ops->status(chip);
   482			if ((status & chip->ops->req_complete_mask) ==
   483			    chip->ops->req_complete_val)
   484				goto out_recv;
   485	
   486			if (chip->ops->req_canceled(chip, status)) {
   487				dev_err(&chip->dev, "Operation Canceled\n");
   488				rc = -ECANCELED;
   489				goto out;
   490			}
   491	
   492			tpm_msleep(TPM_TIMEOUT_POLL);
   493			rmb();
   494		} while (time_before(jiffies, stop));
   495	
   496		chip->ops->cancel(chip);
   497		dev_err(&chip->dev, "Operation Timed out\n");
   498		rc = -ETIME;
   499		goto out;
   500	
   501	out_recv:
   502		len = chip->ops->recv(chip, buf, bufsiz);
   503		if (len < 0) {
   504			rc = len;
   505			dev_err(&chip->dev,
   506				"tpm_transmit: tpm_recv: error %d\n", rc);
   507			goto out;
   508		} else if (len < TPM_HEADER_SIZE) {
   509			rc = -EFAULT;
   510			goto out;
   511		}
   512	
   513		if (len != be32_to_cpu(header->length)) {
   514			rc = -EFAULT;
   515			goto out;
   516		}
   517	
   518		rc = tpm2_commit_space(chip, space, ordinal, buf, &len);
   519	
   520	out:
   521		if (chip->dev.parent)
   522			pm_runtime_put_sync(chip->dev.parent);
   523	
   524		if (need_locality)
   525			tpm_relinquish_locality(chip);
   526	
   527	out_no_locality:
   528		if (chip->ops->clk_enable != NULL)
   529			chip->ops->clk_enable(chip, false);
   530	
   531		if (!(flags & TPM_TRANSMIT_UNLOCKED))
   532			mutex_unlock(&chip->tpm_mutex);
   533		return rc ? rc : len;
   534	}
   535	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH v2] tpm: Return the actual size when receiving an unsupported command
  2018-06-20  9:55 [PATCH v2] tpm: Return the actual size when receiving an unsupported command Ricardo Schwarzmeier
  2018-06-20 14:14 ` kbuild test robot
@ 2018-06-21 16:48 ` Jarkko Sakkinen
  1 sibling, 0 replies; 3+ messages in thread
From: Jarkko Sakkinen @ 2018-06-21 16:48 UTC (permalink / raw)
  To: Ricardo Schwarzmeier
  Cc: linux-integrity, linux-kernel, peterhuewe, jgg, arnd, gregkh

On Wed, Jun 20, 2018 at 11:55:07AM +0200, Ricardo Schwarzmeier wrote:
> The command size is computed for the header anyway.
> It can as well indicate the correct number of bytes to read from the driver.
> Fixes: 095531f891e6 ("tpm: return a TPM_RC_COMMAND_CODE response if command is not implemented")
> 
> Signed-off-by: Ricardo Schwarzmeier <Ricardo.Schwarzmeier@infineon.com>

Can you fix the kbuild error and send v3? Otherwise, LGMT.

/Jarkko

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2018-06-21 16:48 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-06-20  9:55 [PATCH v2] tpm: Return the actual size when receiving an unsupported command Ricardo Schwarzmeier
2018-06-20 14:14 ` kbuild test robot
2018-06-21 16:48 ` Jarkko Sakkinen

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).