From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 7615CC4332F for ; Tue, 31 Oct 2023 08:06:41 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:In-Reply-To:MIME-Version:References: Message-ID:Subject:Cc:To:From:Date:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=X03UBwJDV7zruJiGFThmxcgaaELkDcJox8eGSwVBq9s=; b=Y2acuFBfnQ/Gpg 9m1z9upYnL6jeYRgsEu+/+Wq4L11Ynr7zZ9PflCcIibvvVqvzNFN8pP3MHIvJaYTe4MJslBqKiUY2 QX2MSqI1nsUNR4eGAnbEpvvZHmGe+/3rHSfX4gQ4JGExbo5GOizpgc2m/inrITmEpBKHC0LbaXYxj DZkEw6o6Y/5lKMe72s5rj4+tND7By/0E1kB+WZmi5w9bF8k5ZWhMTBXJkpSKrJ01U7aYomdJwxE/x DVAxaQ9IAARxoVViUDDvfkNddagVv8AfwgIMZuDCJiUrqjOStysFgoe4anZjz+9R2HaUzW9FnDw2b 2Jd0NW5ciV9Tu6P7VnHA==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.96 #2 (Red Hat Linux)) id 1qxjlE-004k3A-05; Tue, 31 Oct 2023 08:06:04 +0000 Received: from foss.arm.com ([217.140.110.172]) by bombadil.infradead.org with esmtp (Exim 4.96 #2 (Red Hat Linux)) id 1qxjl9-004k2m-25 for linux-arm-kernel@lists.infradead.org; Tue, 31 Oct 2023 08:06:02 +0000 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 76076C15; Tue, 31 Oct 2023 01:06:35 -0700 (PDT) Received: from pluto (unknown [172.31.20.19]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id 0224B3F67D; Tue, 31 Oct 2023 01:05:52 -0700 (PDT) Date: Tue, 31 Oct 2023 08:05:50 +0000 From: Cristian Marussi To: "xinglong.yang" Cc: xinglong.yang@cixtech.com, sudeep.holla@arm.com, linux-arm-kernel@lists.infradead.org Subject: Re: [PATCH] firmware: arm_scmi:: Add opinion for init fastchannel Message-ID: References: <20231031013601.925009-1-xinglong.yang@cixtech.com> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <20231031013601.925009-1-xinglong.yang@cixtech.com> X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20231031_010600_662859_96BAE08A X-CRM114-Status: GOOD ( 24.04 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org On Tue, Oct 31, 2023 at 09:36:01AM +0800, xinglong.yang wrote: > Fastchannel may not be supported by the platform. It is not need to > init the fastchannel if the fastchannel is not supported. > Hi, the commit message is misleading because the patch you provided does not really does that :D FCs are indeed not necessarily supported by the platform, and this is reported for each single domain at discovery time AND we indeed check already that (dom->perf_fastchannels) BEFORE initiliazing the fastchannel at all for a specific domain. What you are doing here, instead, is skipping FC initialization on a fastchannel by checking the dom->set_perf / dom->set_limit, BUT such flags are really meant to describe if any set_perf/limit ops can be issued on a domain at all. Given that each FC is associated to a specific domain_id AND msg_id it is indeed plausible that a domain supports FastChannels BUT only for get_* operations. In such scenario I would expect, though, the platform to return NOT_SUPPORTED when required to describe the SET_* FCs. Having said that, it could be anyway useful to refrain from trying to describe/init the SET operations on a FC for a domain where the set_* operations are not supported, because: (1.) avoids unneeded SCMI exchanges, (2.) avoids to trust the FW reply blindly... ...BUT your patch does NOT really do this, because you are skipping the fastchannel_init by looking at dom->set_perf and dom->set_limit for BOTH the set and get commands. You should just skip the FC init on the SET_ when dom->set_ is set if we want to do this optimization/hardening. Am I missing something ? Thanks, Cristian > Signed-off-by: xinglong.yang > Change-Id: Id73ab1f37d5a3726243f97beb40c5b2239d65727 > --- > drivers/firmware/arm_scmi/perf.c | 38 ++++++++++++++++++-------------- > 1 file changed, 21 insertions(+), 17 deletions(-) > > diff --git a/drivers/firmware/arm_scmi/perf.c b/drivers/firmware/arm_scmi/perf.c > index ecf5c4de851b..26fa71e2aff8 100644 > --- a/drivers/firmware/arm_scmi/perf.c > +++ b/drivers/firmware/arm_scmi/perf.c > @@ -526,32 +526,36 @@ static int scmi_perf_level_limits_notify(const struct scmi_protocol_handle *ph, > } > > static void scmi_perf_domain_init_fc(const struct scmi_protocol_handle *ph, > - u32 domain, struct scmi_fc_info **p_fc) > + u32 domain, struct perf_dom_info* dom) > { > + struct scmi_fc_info **p_fc = &dom->fc_info; > struct scmi_fc_info *fc; > > fc = devm_kcalloc(ph->dev, PERF_FC_MAX, sizeof(*fc), GFP_KERNEL); > if (!fc) > return; > > - ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > - PERF_LEVEL_SET, 4, domain, > - &fc[PERF_FC_LEVEL].set_addr, > - &fc[PERF_FC_LEVEL].set_db); > + if (dom->set_perf) { > + ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > + PERF_LEVEL_SET, 4, domain, > + &fc[PERF_FC_LEVEL].set_addr, > + &fc[PERF_FC_LEVEL].set_db); > > - ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > - PERF_LEVEL_GET, 4, domain, > - &fc[PERF_FC_LEVEL].get_addr, NULL); > - > - ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > - PERF_LIMITS_SET, 8, domain, > - &fc[PERF_FC_LIMIT].set_addr, > - &fc[PERF_FC_LIMIT].set_db); > + ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > + PERF_LEVEL_GET, 4, domain, > + &fc[PERF_FC_LEVEL].get_addr, NULL); > + } > > - ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > - PERF_LIMITS_GET, 8, domain, > - &fc[PERF_FC_LIMIT].get_addr, NULL); > + if (dom->set_limits) { > + ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > + PERF_LIMITS_SET, 8, domain, > + &fc[PERF_FC_LIMIT].set_addr, > + &fc[PERF_FC_LIMIT].set_db); > > + ph->hops->fastchannel_init(ph, PERF_DESCRIBE_FASTCHANNEL, > + PERF_LIMITS_GET, 8, domain, > + &fc[PERF_FC_LIMIT].get_addr, NULL); > + } > *p_fc = fc; > } > > @@ -835,7 +839,7 @@ static int scmi_perf_protocol_init(const struct scmi_protocol_handle *ph) > scmi_perf_describe_levels_get(ph, domain, dom); > > if (dom->perf_fastchannels) > - scmi_perf_domain_init_fc(ph, domain, &dom->fc_info); > + scmi_perf_domain_init_fc(ph, domain, dom); > } > > pinfo->version = version; > -- > 2.42.0 > _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel