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 4DB37D44167 for ; Tue, 19 Nov 2024 15:42:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender:List-Subscribe:List-Help :List-Post:List-Archive:List-Unsubscribe:List-Id:Content-Transfer-Encoding: Content-Type:MIME-Version:Message-ID:Date:References:In-Reply-To:Subject:Cc: To:From:Reply-To:Content-ID:Content-Description:Resent-Date:Resent-From: Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner; bh=+KJPrpVSpjlMsljEONMGLhUb9kDFkewDTz7vO3IORSI=; b=JfjVFCn9EtEG+sLJCumS5CY2lF jfJqR7tes2tj44HstTshC49wR3BCWP3vVRYwkkQ65tPfKqFzwS17sASwBVU+po5hFDs+vCsxonVyd dm9TblVd9ZOogq1lE19wGOYzCCIHVjFL7G4fsaxt4VelDssiPWOpG32+faTcQtVaJF5cNWMjfw+OY sYJK3xKyo45PG3csU5vjvdKVoZ9RhVuuY4Ks5ubZtFEnSTun6O33TrnFEGXIJQqN/fvqFjkPopexl i9/YrPWLRlaRyf7V07ApjlH08+YEnIfvo8qKpQ7luoJsMEkKVA//A7aEQ90+F1Ym+NK7NdRbBw8Ed C5v5m0/w==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.98 #2 (Red Hat Linux)) id 1tDQN4-0000000CtqM-3R97; Tue, 19 Nov 2024 15:42:30 +0000 Received: from relay8-d.mail.gandi.net ([2001:4b98:dc4:8::228]) by bombadil.infradead.org with esmtps (Exim 4.98 #2 (Red Hat Linux)) id 1tDQM9-0000000CtfA-0PAi for linux-arm-kernel@lists.infradead.org; Tue, 19 Nov 2024 15:41:34 +0000 Received: by mail.gandi.net (Postfix) with ESMTPSA id DCE0B1BF20C; Tue, 19 Nov 2024 15:41:28 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=bootlin.com; s=gm1; t=1732030889; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=+KJPrpVSpjlMsljEONMGLhUb9kDFkewDTz7vO3IORSI=; b=io9RKrrlUJ5LfS28I+zGTgdXazwkVChYn/JjjVqtb5QYNC9yIXDAhmkzunCCdpOx3BIUOX 0FhAT9sqGU1TAcPeslSlemLGuz4nik2+LXLN7Syk/AHbOpnn7v7qKEC5BGXk600IObMBUJ k5tnn3K6hWPGJO+yM8BXvIE/TDFTToxotzYUcxhavwFJs+BT9FsI3Ea8sJOhRtTpL+rsnA npx/M+CQj9Okjc85fwF1w20qH0V3N2uLAzbBzdL0KwjsgZzQHfpVjaMDCEo6lrUaktYqa0 PuqjD4weYeHzqovr/76QnUG6wc4hvn1b8VxXm8nGwZFrpSm7uBxu3FmF3d2tJA== From: Miquel Raynal To: Marek Vasut Cc: Luca Ceresoli , Abel Vesa , linux-clk@vger.kernel.org, Fabio Estevam , "Lukas F . Hartmann" , Michael Turquette , Peng Fan , Pengutronix Kernel Team , Sascha Hauer , Shawn Guo , Stephen Boyd , imx@lists.linux.dev, kernel@dh-electronics.com, linux-arm-kernel@lists.infradead.org, Anson Huang Subject: Re: [PATCH] clk: imx: clk-imx8mp: Allow media_disp pixel clock reconfigure parent rate In-Reply-To: (Marek Vasut's message of "Mon, 18 Nov 2024 15:30:04 +0100") References: <20240531202648.277078-1-marex@denx.de> <20241112234206.558d5d5e@booty> <79f21303-b0ba-45ed-a842-7e5364fd4efc@denx.de> <20241113120622.3501db73@booty> <130fe140-e70d-4c45-aaab-e22762c58c88@denx.de> <20241115180936.4ab56be3@booty> <6bc5b8d7-ff10-4860-ac46-1460a7d850da@denx.de> <871pz9c606.fsf@bootlin.com> User-Agent: mu4e 1.12.7; emacs 29.4 Date: Tue, 19 Nov 2024 16:41:28 +0100 Message-ID: <87cyirz0wn.fsf@bootlin.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-GND-Sasl: miquel.raynal@bootlin.com X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20241119_074133_421291_79CDBD9A X-CRM114-Status: GOOD ( 31.96 ) 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: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org Hi Marek, On 18/11/2024 at 15:30:04 +01, Marek Vasut wrote: > On 11/18/24 9:15 AM, Miquel Raynal wrote: >> Hi Marek, > > Hello Miquel, > >>>>> If you really want accurate pixel clock for your panel, you need simi= lar >>>>> change to 4fbb73416b10 and configure the Video PLL such that the >>>>> accurate pixel clock can be derived from it. The Video PLL cannot be = set >>>>> to pixel clock, because the LDB serializer clock are either 7x the pi= xel >>>>> clock, or 3.5x the pixel clock (for dual link LVDS), so the Video PLL >>>>> has to be set to 7x or 3.5x pixel clock of the panel, then you should >>>>> get accurate pixel clock and a working panel again. >>>> I found that I'm having the same issue that has been discussed in some >>>> related threads: the lcdif2 configures the video_pll1 to ~72 MHz, and >>>> later LDB tries to set it to 7x that value, failing. >>> >>> Right, which is solved by configuring the Video PLL to the correct >>> frequency in DT up front ... unless you have more than one output >>> supplied by that Video PLL. >> No, this looks like a bug in the imx8 clock driver. I would expect the >> core to handle such case without DT hack. It is not okay to fix clock >> frequencies in DT because drivers are failing to do it properly. I >> understand there are advanced/dual cases with very specific frequencies >> where you don't expect it to magically work and giving hints with DT >> assigned-clocks* properties makes sense, but here I don't think we >> should consider it as a proper fix. > > It is not a proper fix, it is the best we can do right now. I already I am sorry I probably misunderstood your previous reply then. I am fine with the assigned-clocks workaround. > replied to Luca with a bunch of patches where I tried to come up with a > way to negotiate the pixel clock in drivers ... I need to get back to > those. Indeed, thanks to your feedback we got it fixed locally, so short term is okay for us (but people not reading this thread might still suffer from the problem though). >> If I may recap: >> 1- a simple display pipeline works >> 2- the pixel frequency could be more precise so the video_pll1 parent is >> used to dynamically compute a better frequency >> 3- the video_pll1 parent is too low in some cases which breaks the >> pipeline >> 4- we need to force video_pll1 to a value in DT >> How possibly 4 could be a relevant answer to 2, seriously? May I >> return >> you the advice, if you want a better video_pll1 value in the first >> place, why not assigning it up front in DT? > > Because I have DSI-to-(e)DP bridge on the DSI bus and I do not know the > pixel clock needed by attached panel up front. > > I already included a link to DTO which allowed me to operate both this > DSI-to-(e)DP bridge and LVDS panel with accurate pixel clock, I was > hoping that would also let you solve 3 and 4. 4fbb73416b10 ("arm64: dts: > imx8mp-phyboard-pollux: Set Video PLL1 frequency to 506.8 MHz") fixed > 3. for Isaac at least. > >> I understand your goal, and I agree with it, but please acknowledge that >> even though the current patch looks fine per-se, it is exposing a real >> bug that is now visible. Hiding it with DT properties feels really wrong. > I do fully agree the whole DT Video PLL1 clock frequency configuration > is not good and it should not be in the DT at all. That is my goal in > the very end. > > The drivers (in this case, LCDIF1 + LCDIF2 + LDB) should negotiate the > Video PLL1 frequency that fits them all best and configure it > accordingly, without any DT assign-clock* workarounds. Ok, good to know we are aligned :-) > I just didn't figure out a way to do that ^ yet. Of course, getting rid of the DT workarounds is probably a long term goal, unlike the mid-term goal which is: "fixing" today's situation for "everyone with a simple setup". We are also looking into this and willing to find a proper solution. Cheers, Miqu=C3=A8l