From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtpout-03.galae.net (smtpout-03.galae.net [185.246.85.4]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 04A733B5301; Mon, 9 Mar 2026 12:54:37 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=185.246.85.4 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1773060880; cv=none; b=u9kC9Ldv0Zw3l3UPNkh6K1kui1+c5Qz6XD3UtXyN4Mby8W7G9afUnDwJHSmBw4dfGIsuNEqc4YPASJKkSf7o8y9H4NXS4b+FCQulLxDOBY3jVsLMkBuAsQFPoJ5LUeJbLjxQ4Chk3RLfe786y5qpKLT/OPNm4P5tp/KZbVNWAgY= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1773060880; c=relaxed/simple; bh=JEs2IWfmscrKw+t0H8dNptKKC8kR/QKpS4L+iYAFxzc=; h=Message-ID:Date:MIME-Version:Subject:To:Cc:References:From: In-Reply-To:Content-Type; b=mCgCT0cLHyVb0iWnx/WeXO5j+GrpMLTLZphQDq6qI7oDAPDYEJO2izaezNsXzu9ZGh7dwZYFK67yTmF7abrmY1jBv5iVVSwCt54FpF3ttdTcrQGKMvpRWfhpAQfmjwm8ZS4ctHUCU8lgKsefJsONK48ECjWyGSKNLkJs30iyHkM= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=bootlin.com; spf=pass smtp.mailfrom=bootlin.com; dkim=pass (2048-bit key) header.d=bootlin.com header.i=@bootlin.com header.b=KnE2JS/Z; arc=none smtp.client-ip=185.246.85.4 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=bootlin.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=bootlin.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=bootlin.com header.i=@bootlin.com header.b="KnE2JS/Z" Received: from smtpout-01.galae.net (smtpout-01.galae.net [212.83.139.233]) by smtpout-03.galae.net (Postfix) with ESMTPS id 50C9E4E425D9; Mon, 9 Mar 2026 12:54:36 +0000 (UTC) Received: from mail.galae.net (mail.galae.net [212.83.136.155]) by smtpout-01.galae.net (Postfix) with ESMTPS id 231675FFB8; Mon, 9 Mar 2026 12:54:36 +0000 (UTC) Received: from [127.0.0.1] (localhost [127.0.0.1]) by localhost (Mailerdaemon) with ESMTPSA id CD71D10369B92; Mon, 9 Mar 2026 13:54:29 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=bootlin.com; s=dkim; t=1773060874; h=from:subject:date:message-id:to:cc:mime-version:content-type: content-transfer-encoding:content-language:in-reply-to:references; bh=ZR9GAoP8GpYeY6MkEWjSdSANaRmKWabJF46IrFRn1Rs=; b=KnE2JS/ZI7dVsvlB0zYO641VEYu0fTPBI6o2A4cgrTRwMEN2aU5gDTdAjNB0+W86lC1F1l BGmz5tSb8p2zExVVHVpSCD3XgdSOB30Z6E15x3qxTMM361tk4IbQ3QZPccTfzzXtXV9X4M i+VO61uF9NVrGBHsQQa/SAI4By/eVagw39Vq5phVrNa4K4fO/TgwgoVdO5wIzUXlyl52nE iHgg5yNFnCFAaxpvLc0Dc33EMFdb850TR2zIZCwMS+T+yCMLSCUcgLduvWs1uWYY/0hdIL 8ySEFnz6RucXDfeDMHEw5NBhZZYQEShfGdlPaA6rsKLGkdYrnZi7W2x77bL76g== Message-ID: <0658868d-671c-46a2-8daa-26420cdd2975@bootlin.com> Date: Mon, 9 Mar 2026 13:54:28 +0100 Precedence: bulk X-Mailing-List: netdev@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH net-next v6 1/9] net: dsa: microchip: Add support for KSZ8463 global irq To: olteanv@gmail.com Cc: Woojung.Huh@microchip.com, UNGLinuxDriver@microchip.com, andrew@lunn.ch, davem@davemloft.net, edumazet@google.com, kuba@kernel.org, pabeni@redhat.com, richardcochran@gmail.com, horms@kernel.org, pascal.eberhard@se.com, miquel.raynal@bootlin.com, thomas.petazzoni@bootlin.com, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, maxime.chevallier@bootlin.com, Tristram.Ha@microchip.com References: <20260304-ksz8463-ptp-v6-0-3f4c47954c71@bootlin.com> <20260304-ksz8463-ptp-v6-1-3f4c47954c71@bootlin.com> <20260305095656.vlyaztv6nbdqrmil@skbuf> <98944cef-0877-4fb9-83a0-92bbd3852f66@bootlin.com> <20260305125149.ejju5ptrkviqi3sm@skbuf> From: Bastien Curutchet Content-Language: en-US In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-Last-TLS-Session-Version: TLSv1.3 Hi Vladimir, On 3/6/26 2:10 AM, Tristram.Ha@microchip.com wrote: >> On 3/5/26 1:51 PM, Vladimir Oltean wrote: >>> On Thu, Mar 05, 2026 at 01:39:29PM +0100, Bastien Curutchet wrote: >>>> Hi Vladimir, >>>> >>>> On 3/5/26 10:56 AM, Vladimir Oltean wrote: >>>>> On Wed, Mar 04, 2026 at 11:18:52AM +0100, Bastien Curutchet (Schneider >> Electric) wrote: >>>>>> @@ -2890,14 +2899,18 @@ static irqreturn_t ksz_irq_thread_fn(int irq, >> void *dev_id) >>>>>> unsigned int nhandled = 0; >>>>>> struct ksz_device *dev; >>>>>> unsigned int sub_irq; >>>>>> - u8 data; >>>>>> + u16 data; >>>>>> int ret; >>>>>> u8 n; >>>>>> dev = kirq->dev; >>>>>> - /* Read interrupt status register */ >>>>>> - ret = ksz_read8(dev, kirq->reg_status, &data); >>>>>> + /* >>>>>> + * Most of the KSZ switches have a 8-bits long interrupt status >>>>>> + * register, but the KSZ8463 has a 16-bits long one. The overread here >>>>>> + * is safe because we only iterate over kirq->nirqs in the below loop. >>>>> >>>>> FWIW, this isn't the only thing making an overread "safe". >>>>> If the adjacent register also has "clear on read" semantics, that's not >>>>> good. >>>>> >>>>> I can't tell whether that's the case here, though. There are just too >>>>> many hardware variations to check for. I just wanted to point out that >>>>> the reasoning is incomplete. >>>>> >>>> You're right, I hadn't thought about the 'clear on read' case. >>>> >>>> I'll use ksz_read16() only for the KSZ8463 to be 100% sure it won't break >>>> anything for other switches. >>> >>> I'm still on the fence on whether to say this or not, because I don't >>> really want to get so involved in internal driver bookkeeping, but... >>> this driver is just becoming a hell to review, even if I want to >>> concentrate exclusively on correct API use, like I try to do. >>>> Linus Walleij has added a new ks8995 driver, which has some overlap with >>> the common ksz driver for the KSZ8 family. Now he wants to remove the >>> overlapping device support: >>> https://lore.kernel.org/netdev/20260219-ks8995-fixups-v3-0- >> a7fc63fe1916@kernel.org/ >>> >>> Maybe we should go the other way around, migrate KSZ8 support to the >>> ks8995 driver instead? The common ksz driver is becoming just extremely >>> convoluted to handle all hardware variations. Would it help in any way >>> to maintain cleaner code paths, what do you think? >> >> I agree, this driver is extremely convoluted because of all the >> different hardware it handles. It wasn't easy to fit PTP support for the >> KSZ8463 into it. And I encountered the same kind of difficulties when >> adding periodic output support (I have another series ready for this >> once this PTP support will be merged). >> >> Regarding migrating the KSZ8 support into the ksz8995, I think we would >> quickly hit the same pain points than in ksz_common. Even inside the >> KSZ8 family we can find a quite big amount of differences between the >> switches. For instance, both KSZ8463 and KSZ8563 support PTP, they share >> lots of common registers but their interrupt scheme is very different. >> >> I've added Tristram in Cc:, who works at Microchip. Maybe, Tristram, you >> have some insights about which switches could share code if we decide to >> split the big ksz_common into several drivers ? > > Although KSZ8463 uses 16-bit register for its interrupt operation, the > bits that are necessary for actual operation are all in the high byte, so > it is possible to simulate using 8-bit register for interrupt operation > just like the other switches. > > Note also the previous interrupt code only works for KSZ9477 and LAN937X > families of switches, which have interrupts for each port. The older > switches like KSZ8863, KSZ8895, and KSZ8795 all have only one global > interrupt and need change to work in this framework. > > These switches have link change interrupt bits for external ports, so it > is easy to simulate them as port interrupts. But KSZ8463 only has 1 bit > for link, so it is necessary to add some special code to return a value of > 3 to indicate both ports have link change. > > KSZ8463 is based from KSZ8863 so it should have about the same 8-bit > register definitions as KSZ8863. But when it was designed it had a > companion chip that operates as a 16-bit network controller. Because of > that all register definitions are described in 16-bit. > > I need to try your patches to see how it works regarding to the PTP > interrupts. > > Regarding to the old ks8995 driver I think it was a simple SPI driver to > start the KSZ8995 switch, which is an older version of KSZ8895. It was > used to locate under the PHY drivers. I did not know it was upgraded to > become a full DSA driver and is now trying to expand to the other > switches. > I have a new iteration ready. It uses only the high byte of the interrupt registers for the KSZ8463, which keeps unchanged the current 8-bit accesses logic. Shall I send it ? Or would you like more time to think/discuss about whether we should split ksz_common into several drivers ? Best regards, Bastien