| Message ID | 20171017203232.2262-4-nayna@linux.vnet.ibm.com (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
On Tue, Oct 17, 2017 at 04:32:31PM -0400, Nayna Jain wrote: > The existing wait_for_tpm_stat() polls for the chip status after > 5msec sleep. As per TCG ddwg input, it is expected that tpm might > return status in few usec. So, reducing the delay in polling to > 1msec. > Similarly, get_burstcount() function sleeps for 5msec before > retrying for next query to burstcount in a loop. If it takes > lesser time for TPM to return, this 5msec delay is longer than > necessary. > > After this change, performance on a TPM 1.2 with an 8 byte > burstcount for 1000 extends improved from ~14sec to ~9sec. > > Signed-off-by: Nayna Jain <nayna@linux.vnet.ibm.com> Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com> /Jarkko > --- > drivers/char/tpm/tpm_tis_core.c | 10 ++++++++-- > 1 file changed, 8 insertions(+), 2 deletions(-) > > diff --git a/drivers/char/tpm/tpm_tis_core.c b/drivers/char/tpm/tpm_tis_core.c > index 993328ae988c..8b55fdaa67f2 100644 > --- a/drivers/char/tpm/tpm_tis_core.c > +++ b/drivers/char/tpm/tpm_tis_core.c > @@ -31,6 +31,12 @@ > #include "tpm.h" > #include "tpm_tis_core.h" > > +/* This is a polling delay to check for status and burstcount. > + * As per ddwg input, expectation is that status check and burstcount > + * check should return within few usecs. > + */ > +#define TPM_POLL_SLEEP 1 /* msec */ > + > static bool wait_for_tpm_stat_cond(struct tpm_chip *chip, u8 mask, > bool check_cancel, bool *canceled) > { > @@ -82,7 +88,7 @@ static int wait_for_tpm_stat(struct tpm_chip *chip, u8 mask, > } > } else { > do { > - tpm_msleep(TPM_TIMEOUT); > + tpm_msleep(TPM_POLL_SLEEP); > status = chip->ops->status(chip); > if ((status & mask) == mask) > return 0; > @@ -224,7 +230,7 @@ static int get_burstcount(struct tpm_chip *chip) > burstcnt = (value >> 8) & 0xFFFF; > if (burstcnt) > return burstcnt; > - tpm_msleep(TPM_TIMEOUT); > + tpm_msleep(TPM_POLL_SLEEP); > } while (time_before(jiffies, stop)); > return -EBUSY; > } > -- > 2.13.3 >
On Wed, Oct 18, 2017 at 06:24:47PM +0300, Jarkko Sakkinen wrote: > On Tue, Oct 17, 2017 at 04:32:31PM -0400, Nayna Jain wrote: > > The existing wait_for_tpm_stat() polls for the chip status after > > 5msec sleep. As per TCG ddwg input, it is expected that tpm might > > return status in few usec. So, reducing the delay in polling to > > 1msec. > > Similarly, get_burstcount() function sleeps for 5msec before > > retrying for next query to burstcount in a loop. If it takes > > lesser time for TPM to return, this 5msec delay is longer than > > necessary. > > > > After this change, performance on a TPM 1.2 with an 8 byte > > burstcount for 1000 extends improved from ~14sec to ~9sec. > > > > Signed-off-by: Nayna Jain <nayna@linux.vnet.ibm.com> > > Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com> Tested-by: Jarkko Sakkinen <jarkkko.sakkinen@linux.intel.com> /Jarkko
diff --git a/drivers/char/tpm/tpm_tis_core.c b/drivers/char/tpm/tpm_tis_core.c index 993328ae988c..8b55fdaa67f2 100644 --- a/drivers/char/tpm/tpm_tis_core.c +++ b/drivers/char/tpm/tpm_tis_core.c @@ -31,6 +31,12 @@ #include "tpm.h" #include "tpm_tis_core.h" +/* This is a polling delay to check for status and burstcount. + * As per ddwg input, expectation is that status check and burstcount + * check should return within few usecs. + */ +#define TPM_POLL_SLEEP 1 /* msec */ + static bool wait_for_tpm_stat_cond(struct tpm_chip *chip, u8 mask, bool check_cancel, bool *canceled) { @@ -82,7 +88,7 @@ static int wait_for_tpm_stat(struct tpm_chip *chip, u8 mask, } } else { do { - tpm_msleep(TPM_TIMEOUT); + tpm_msleep(TPM_POLL_SLEEP); status = chip->ops->status(chip); if ((status & mask) == mask) return 0; @@ -224,7 +230,7 @@ static int get_burstcount(struct tpm_chip *chip) burstcnt = (value >> 8) & 0xFFFF; if (burstcnt) return burstcnt; - tpm_msleep(TPM_TIMEOUT); + tpm_msleep(TPM_POLL_SLEEP); } while (time_before(jiffies, stop)); return -EBUSY; }
The existing wait_for_tpm_stat() polls for the chip status after 5msec sleep. As per TCG ddwg input, it is expected that tpm might return status in few usec. So, reducing the delay in polling to 1msec. Similarly, get_burstcount() function sleeps for 5msec before retrying for next query to burstcount in a loop. If it takes lesser time for TPM to return, this 5msec delay is longer than necessary. After this change, performance on a TPM 1.2 with an 8 byte burstcount for 1000 extends improved from ~14sec to ~9sec. Signed-off-by: Nayna Jain <nayna@linux.vnet.ibm.com> --- drivers/char/tpm/tpm_tis_core.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)