Date:   Thu, 5 Mar 2020 17:00:26 +0530
From:   Maulik Shah <mkshah@...eaurora.org>
To:     Doug Anderson <dianders@...omium.org>
Cc:     Stephen Boyd <swboyd@...omium.org>,
        Matthias Kaehlcke <mka@...omium.org>,
        Evan Green <evgreen@...omium.org>,
        Bjorn Andersson <bjorn.andersson@...aro.org>,
        LKML <linux-kernel@...r.kernel.org>,
        linux-arm-msm <linux-arm-msm@...r.kernel.org>,
        Andy Gross <agross@...nel.org>,
        Rajendra Nayak <rnayak@...eaurora.org>,
        Lina Iyer <ilina@...eaurora.org>, lsrao@...eaurora.org
Subject: Re: [PATCH v10 3/3] soc: qcom: rpmh: Invoke rpmh_flush() for dirty
 caches


On 3/5/2020 4:52 AM, Doug Anderson wrote:
> Hi,
>
> On Tue, Mar 3, 2020 at 4:27 AM Maulik Shah <mkshah@...eaurora.org> wrote:
>> Add changes to invoke rpmh flush() from within cache_lock when the data
>> in cache is dirty.
>>
>> This is done only if OSI is not supported in PSCI. If OSI is supported
>> rpmh_flush can get invoked when the last cpu going to power collapse
>> deepest low power mode.
>>
>> Also remove "depends on COMPILE_TEST" for Kconfig option QCOM_RPMH so the
>> driver is only compiled for arm64 which supports psci_has_osi_support()
>> API.
>>
>> Signed-off-by: Maulik Shah <mkshah@...eaurora.org>
>> Reviewed-by: Srinivas Rao L <lsrao@...eaurora.org>
>> ---
>>  drivers/soc/qcom/Kconfig |  2 +-
>>  drivers/soc/qcom/rpmh.c  | 37 ++++++++++++++++++++++---------------
>>  2 files changed, 23 insertions(+), 16 deletions(-)
>>
>> diff --git a/drivers/soc/qcom/Kconfig b/drivers/soc/qcom/Kconfig
>> index d0a73e7..2e581bc 100644
>> --- a/drivers/soc/qcom/Kconfig
>> +++ b/drivers/soc/qcom/Kconfig
>> @@ -105,7 +105,7 @@ config QCOM_RMTFS_MEM
>>
>>  config QCOM_RPMH
>>         bool "Qualcomm RPM-Hardened (RPMH) Communication"
>> -       depends on ARCH_QCOM && ARM64 || COMPILE_TEST
>> +       depends on ARCH_QCOM && ARM64
>>         help
>>           Support for communication with the hardened-RPM blocks in
>>           Qualcomm Technologies Inc (QTI) SoCs. RPMH communication uses an
>> diff --git a/drivers/soc/qcom/rpmh.c b/drivers/soc/qcom/rpmh.c
>> index f28afe4..dafb0da 100644
>> --- a/drivers/soc/qcom/rpmh.c
>> +++ b/drivers/soc/qcom/rpmh.c
>> @@ -12,6 +12,7 @@
>>  #include <linux/module.h>
>>  #include <linux/of.h>
>>  #include <linux/platform_device.h>
>> +#include <linux/psci.h>
>>  #include <linux/slab.h>
>>  #include <linux/spinlock.h>
>>  #include <linux/types.h>
>> @@ -158,6 +159,13 @@ static struct cache_req *cache_rpm_request(struct rpmh_ctrlr *ctrlr,
>>         }
>>
>>  unlock:
>> +       if (ctrlr->dirty && !psci_has_osi_support()) {
>> +               if (rpmh_flush(ctrlr)) {
>> +                       spin_unlock_irqrestore(&ctrlr->cache_lock, flags);
>> +                       return ERR_PTR(-EINVAL);
>> +               }
>> +       }
>> +
>>         spin_unlock_irqrestore(&ctrlr->cache_lock, flags);
>>
>>         return req;
>> @@ -285,26 +293,35 @@ int rpmh_write(const struct device *dev, enum rpmh_state state,
>>  }
>>  EXPORT_SYMBOL(rpmh_write);
>>
>> -static void cache_batch(struct rpmh_ctrlr *ctrlr, struct batch_cache_req *req)
>> +static int cache_batch(struct rpmh_ctrlr *ctrlr, struct batch_cache_req *req)
>>  {
>>         unsigned long flags;
>>
>>         spin_lock_irqsave(&ctrlr->cache_lock, flags);
>> +
>>         list_add_tail(&req->list, &ctrlr->batch_cache);
>>         ctrlr->dirty = true;
>> +
>> +       if (!psci_has_osi_support()) {
>> +               if (rpmh_flush(ctrlr)) {
> The whole API here is a bit unfortunate.  From what I can tell,
> callers of this code almost always call rpmh_write_batch() in
> triplicate, AKA:
>
> rpmh_write_batch(active, ...)
> rpmh_write_batch(wake, ...)
> rpmh_write_batch(sleep, ...)
>
> ...that's going to end up writing the whole sleep/wake sets twice
> every single time, right?  I know you talked about trying to keep
> separate dirty bits for sleep/wake and maybe that would help, but it
> might not be so easy due to the comparison of "sleep_val" and
> "wake_val" in is_req_valid().
>
> I guess we can keep the inefficiency for now and see how much it hits
> us, but it feels ugly.
>
>
>> +                       spin_unlock_irqrestore(&ctrlr->cache_lock, flags);
>> +                       return -EINVAL;
> nit: why not add "int ret = 0" to the top of the function, then here:
>
> if (rpmh_flush(ctrl))
>   ret = -EINVAL;
>
> ...then at the end "return ret".  It avoids the 2nd copy of the unlock?
Done.
>
> Also: Why throw away the return value of rpmh_flush and replace it
> with -EINVAL?  Trying to avoid -EBUSY?  ...oh, should you handle
> -EBUSY?  AKA:
>
> if (!psci_has_osi_support()) {
>   do {
>     ret = rpmh_flush(ctrl);
>   } while (ret == -EBUSY);
> }

Done, the return value from rpmh_flush() can be -EAGAIN, not -EBUSY.

i will update the comment accordingly and will include below change as well in next series.

https://patchwork.kernel.org/patch/11364067/

this should address for caller to not handle -EAGAIN.

>
>
>> +               }
>> +       }
>> +
>>         spin_unlock_irqrestore(&ctrlr->cache_lock, flags);
>> +
>> +       return 0;
>>  }
>>
>>  static int flush_batch(struct rpmh_ctrlr *ctrlr)
>>  {
>>         struct batch_cache_req *req;
>>         const struct rpmh_request *rpm_msg;
>> -       unsigned long flags;
>>         int ret = 0;
>>         int i;
>>
>>         /* Send Sleep/Wake requests to the controller, expect no response */
>> -       spin_lock_irqsave(&ctrlr->cache_lock, flags);
>>         list_for_each_entry(req, &ctrlr->batch_cache, list) {
>>                 for (i = 0; i < req->count; i++) {
>>                         rpm_msg = req->rpm_msgs + i;
>> @@ -314,7 +331,6 @@ static int flush_batch(struct rpmh_ctrlr *ctrlr)
>>                                 break;
>>                 }
>>         }
>> -       spin_unlock_irqrestore(&ctrlr->cache_lock, flags);
>>
>>         return ret;
>>  }
>> @@ -386,10 +402,8 @@ int rpmh_write_batch(const struct device *dev, enum rpmh_state state,
>>                 cmd += n[i];
>>         }
>>
>> -       if (state != RPMH_ACTIVE_ONLY_STATE) {
>> -               cache_batch(ctrlr, req);
>> -               return 0;
>> -       }
>> +       if (state != RPMH_ACTIVE_ONLY_STATE)
>> +               return cache_batch(ctrlr, req);
>>
>>         for (i = 0; i < count; i++) {
>>                 struct completion *compl = &compls[i];
>> @@ -455,9 +469,6 @@ static int send_single(struct rpmh_ctrlr *ctrlr, enum rpmh_state state,
>>   * Return: -EBUSY if the controller is busy, probably waiting on a response
>>   * to a RPMH request sent earlier.
>>   *
>> - * This function is always called from the sleep code from the last CPU
>> - * that is powering down the entire system. Since no other RPMH API would be
>> - * executing at this time, it is safe to run lockless.
> nit: you've now got an extra "blank" (just has a "*" on it) line at
> the end of your comment block.
Done.
> nit: in v9, Evan suggested "We should probably replace that with a
> comment indicating that we assume ctrlr->cache_lock is already held".
> Maybe you could do that?
yes i left it for below reason since we still can call it from sleep code.
i will mention same in v11.

Thanks,
Maulik
>
> Also: presumably you _will_ still be called by the sleep code from the
> last CPU on systems with OSI.  Is that true?  If that's not true then
> you should change your function to static.  If that is true, then your
> comment should be something like "this function will either be called
> from sleep code on the last CPU (thus no spinlock needed) or with the
> spinlock already held".
>
>
> -Doug

-- 
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation

Powered by blists - more mailing lists