On Wed, Jun 13 2018 at 15:09 -0600, Doug Anderson wrote:
Hi, On Mon, Jun 11, 2018 at 10:25 AM, Raju P L S S S N <rplsssn@xxxxxxxxxxxxxx> wrote:@@ -148,7 +148,8 @@ int rpmh_rsc_invalidate(struct rsc_drv *drv) static struct tcs_group *get_tcs_for_msg(struct rsc_drv *drv, const struct tcs_request *msg) { - int type; + int type, ret; + struct tcs_group *tcs; switch (msg->state) { case RPMH_ACTIVE_ONLY_STATE: @@ -164,7 +165,25 @@ static struct tcs_group *get_tcs_for_msg(struct rsc_drv *drv, return ERR_PTR(-EINVAL); } - return get_tcs_of_type(drv, type); + /* + * If we are making an active request on a RSC that does not have a + * dedicated TCS for active state use, then re-purpose a wake TCS to + * send active votes. + * NOTE: The driver must be aware that this RSC does not have a + * dedicated AMC, and therefore would invalidate the sleep and wake + * TCSes before making an active state request. + */ + tcs = get_tcs_of_type(drv, type); + if (msg->state == RPMH_ACTIVE_ONLY_STATE && IS_ERR(tcs)) { + tcs = get_tcs_of_type(drv, WAKE_TCS); + if (!IS_ERR(tcs)) { + ret = rpmh_rsc_invalidate(drv); + if (ret) + return ERR_PTR(ret);In v9 you looped as long as the "ret" was -EAGAIN. Now you're not. Are all the callers setup to handle -EAGAIN or should you keep the loop in for -EAGAIN? I don't think callers handle this well. ...or is there some reason that EAGAIN can't happen in this call to rpmh_rsc_invalidate()?
Returning an error back to the caller is the right approach here. If the invalidate fails, its mostly because, the TCS is busy or in use by an external entity. Only the caller would be able to resolve the condition and take corrective action. -- Lina -- To unsubscribe from this list: send the line "unsubscribe linux-soc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
