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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 45E91C433F5 for ; Fri, 28 Jan 2022 17:52:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1343501AbiA1Rwt convert rfc822-to-8bit (ORCPT ); Fri, 28 Jan 2022 12:52:49 -0500 Received: from frasgout.his.huawei.com ([185.176.79.56]:4553 "EHLO frasgout.his.huawei.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S243013AbiA1Rwt (ORCPT ); Fri, 28 Jan 2022 12:52:49 -0500 Received: from fraeml707-chm.china.huawei.com (unknown [172.18.147.207]) by frasgout.his.huawei.com (SkyGuard) with ESMTP id 4JllLF0ls6z67brf; Sat, 29 Jan 2022 01:48:21 +0800 (CST) Received: from lhreml710-chm.china.huawei.com (10.201.108.61) by fraeml707-chm.china.huawei.com (10.206.15.35) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2308.21; Fri, 28 Jan 2022 18:52:46 +0100 Received: from localhost (10.122.247.231) by lhreml710-chm.china.huawei.com (10.201.108.61) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2308.21; Fri, 28 Jan 2022 17:52:46 +0000 Date: Fri, 28 Jan 2022 17:52:45 +0000 From: Jonathan Cameron To: Alex =?ISO-8859-1?Q?Benn=E9e?= CC: , Marcel Apfelbaum , "Michael S . Tsirkin" , Igor Mammedov , , Ben Widawsky , "Peter Maydell" , , "Shameerali Kolothum Thodi" , Philippe =?ISO-8859-1?Q?Mathieu-Daud=E9?= , Saransh Gupta1 , Shreyas Shah , Chris Browy , Samarth Saxena , "Dan Williams" Subject: Re: [PATCH v4 09/42] hw/cxl/device: Timestamp implementation (8.2.9.3) Message-ID: <20220128175245.00007f6b@huawei.com> In-Reply-To: <874k5pbdlh.fsf@linaro.org> References: <20220124171705.10432-1-Jonathan.Cameron@huawei.com> <20220124171705.10432-10-Jonathan.Cameron@huawei.com> <874k5pbdlh.fsf@linaro.org> Organization: Huawei Technologies R&D (UK) Ltd. X-Mailer: Claws Mail 4.0.0 (GTK+ 3.24.29; x86_64-w64-mingw32) MIME-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 8BIT X-Originating-IP: [10.122.247.231] X-ClientProxiedBy: lhreml740-chm.china.huawei.com (10.201.108.190) To lhreml710-chm.china.huawei.com (10.201.108.61) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-cxl@vger.kernel.org On Thu, 27 Jan 2022 11:50:08 +0000 Alex Bennée wrote: > Jonathan Cameron writes: > > > From: Ben Widawsky > > > > Per spec, timestamp appears to be a free-running counter from a value > > set by the host via the Set Timestamp command (0301h). There are > > references to the epoch, which seem like a red herring. Therefore, the > > implementation implements the timestamp as freerunning counter from the > > last value that was issued by the Set Timestamp command. > > > > Signed-off-by: Ben Widawsky > > Signed-off-by: Jonathan Cameron > > --- > > hw/cxl/cxl-mailbox-utils.c | 53 +++++++++++++++++++++++++++++++++++++ > > include/hw/cxl/cxl_device.h | 6 +++++ > > 2 files changed, 59 insertions(+) > > > > diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c > > index 1a87846356..cea4b2a59c 100644 > > --- a/hw/cxl/cxl-mailbox-utils.c > > +++ b/hw/cxl/cxl-mailbox-utils.c > > @@ -43,6 +43,9 @@ enum { > > #define CLEAR_RECORDS 0x1 > > #define GET_INTERRUPT_POLICY 0x2 > > #define SET_INTERRUPT_POLICY 0x3 > > + TIMESTAMP = 0x03, > > + #define GET 0x0 > > + #define SET 0x1 > > }; > > > > /* 8.2.8.4.5.1 Command Return Codes */ > > @@ -117,8 +120,11 @@ define_mailbox_handler_zeroed(EVENTS_GET_RECORDS, 0x20); > > define_mailbox_handler_nop(EVENTS_CLEAR_RECORDS); > > define_mailbox_handler_zeroed(EVENTS_GET_INTERRUPT_POLICY, 4); > > define_mailbox_handler_nop(EVENTS_SET_INTERRUPT_POLICY); > > +declare_mailbox_handler(TIMESTAMP_GET); > > +declare_mailbox_handler(TIMESTAMP_SET); > > > > #define IMMEDIATE_CONFIG_CHANGE (1 << 1) > > +#define IMMEDIATE_POLICY_CHANGE (1 << 3) > > #define IMMEDIATE_LOG_CHANGE (1 << 4) > > > > #define CXL_CMD(s, c, in, cel_effect) \ > > @@ -129,10 +135,57 @@ static struct cxl_cmd cxl_cmd_set[256][256] = { > > CXL_CMD(EVENTS, CLEAR_RECORDS, ~0, IMMEDIATE_LOG_CHANGE), > > CXL_CMD(EVENTS, GET_INTERRUPT_POLICY, 0, 0), > > CXL_CMD(EVENTS, SET_INTERRUPT_POLICY, 4, IMMEDIATE_CONFIG_CHANGE), > > + CXL_CMD(TIMESTAMP, GET, 0, 0), > > + CXL_CMD(TIMESTAMP, SET, 8, IMMEDIATE_POLICY_CHANGE), > > }; > > > > #undef CXL_CMD > > > > +/* > > + * 8.2.9.3.1 > > + */ > > +define_mailbox_handler(TIMESTAMP_GET) > > +{ > > + struct timespec ts; > > + uint64_t delta; > > + > > + if (!cxl_dstate->timestamp.set) { > > + *(uint64_t *)cmd->payload = 0; > > + goto done; > > + } > > + > > + /* First find the delta from the last time the host set the time. */ > > + clock_gettime(CLOCK_REALTIME, &ts); > > Could you consider using qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL)? > Otherwise by introducing a dependency on real time you'll loose the > ability to get deterministic execution via icount. I don't have a strong opinion either way, so Ill go with qemu_clock_get_ns() and see if any comments on it in v5. I've also updated the commit message as the catch all errata which was F4 included a clarification that the timestamp was as implemented here. Thanks, Jonathan