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 09F7FCCFA13 for ; Mon, 10 Nov 2025 16:28:27 +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:In-Reply-To:From:References:Cc:To:Subject:MIME-Version:Date: Message-ID:Reply-To:Content-ID:Content-Description:Resent-Date:Resent-From: Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner; bh=z80pbkP/TYorjbm1oWzO0yeXlK/iKwPApxaIQQlIuuE=; b=dWjMk8dFCkQHIWALE5z1eLQSWM G+3nB0pjheK76KGjinNKxBntyFPPbB75t9CsxWssfGYdTFWqce1TojF3DqK+LVeVC/W5SwjskD3pQ imiDs+AH5c0QQ361i+SLJ7pN69eUIvHYDgTtXsFsHJ5QfEnhxqSNM04YKT4k1dt6nzzyYUboIETDT B8GW4oi9NdmamdqEPaWc+5MG9kH+rKZrAJhtLjou30wyk3Vol6139jiuOcEKJ4VlCkdoL2cZx20Pz H6kW4NS2lJ3cLsRpvrj3WGn3m8Flr2jXA9vd/7d9P6CT2Z/cF2gRm5nMuAgJYkCUtZpbO0T6F9uqw BsHaQpYQ==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.98.2 #2 (Red Hat Linux)) id 1vIUkc-00000005nBM-2zsu; Mon, 10 Nov 2025 16:28:18 +0000 Received: from foss.arm.com ([217.140.110.172]) by bombadil.infradead.org with esmtp (Exim 4.98.2 #2 (Red Hat Linux)) id 1vIUka-00000005nAu-3dx1 for linux-arm-kernel@lists.infradead.org; Mon, 10 Nov 2025 16:28:18 +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 5AEA82B; Mon, 10 Nov 2025 08:28:08 -0800 (PST) Received: from [10.1.196.46] (e134344.arm.com [10.1.196.46]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id 7C5463F66E; Mon, 10 Nov 2025 08:28:11 -0800 (PST) Message-ID: <17caf75c-a00f-41d8-bacc-af5ba6c485d9@arm.com> Date: Mon, 10 Nov 2025 16:28:10 +0000 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH 03/33] ACPI / PPTT: Add acpi_pptt_cache_v1_full to use pptt cache as one structure To: Jonathan Cameron Cc: james.morse@arm.com, amitsinght@marvell.com, baisheng.gao@unisoc.com, baolin.wang@linux.alibaba.com, bobo.shaobowang@huawei.com, carl@os.amperecomputing.com, catalin.marinas@arm.com, dakr@kernel.org, dave.martin@arm.com, david@redhat.com, dfustini@baylibre.com, fenghuay@nvidia.com, gregkh@linuxfoundation.org, gshan@redhat.com, guohanjun@huawei.com, jeremy.linton@arm.com, kobak@nvidia.com, lcherian@marvell.com, lenb@kernel.org, linux-acpi@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, lpieralisi@kernel.org, peternewman@google.com, quic_jiles@quicinc.com, rafael@kernel.org, robh@kernel.org, rohit.mathew@arm.com, scott@os.amperecomputing.com, sdonthineni@nvidia.com, sudeep.holla@arm.com, tan.shaopeng@fujitsu.com, will@kernel.org, xhao@linux.alibaba.com References: <20251107123450.664001-1-ben.horgan@arm.com> <20251107123450.664001-4-ben.horgan@arm.com> <20251110154610.00002247@huawei.com> From: Ben Horgan Content-Language: en-US In-Reply-To: <20251110154610.00002247@huawei.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20251110_082817_000791_BAEEF855 X-CRM114-Status: GOOD ( 30.91 ) 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 Jonathan, On 11/10/25 15:46, Jonathan Cameron wrote: > On Fri, 7 Nov 2025 12:34:20 +0000 > Ben Horgan wrote: > >> In actbl2.h, struct acpi_pptt_cache describes the fields in the original >> cache type structure. In PPTT table version 3 a new field was added at the >> end, cache_id. This is described in struct acpi_pptt_cache_v1. Introduce >> the new, acpi_pptt_cache_v1_full to contain both these structures. Update >> the existing code to use this new struct. This simplifies the code, removes >> a non-standard use of ACPI_ADD_PTR and allows using the length in the >> header to check if the cache_id is valid. >> >> Signed-off-by: Ben Horgan > > Whilst I wish the ACPICA stuff did structures like this, I'm not sure > if the ACPI maintainers will feel it is appropriate to work around it > with generic sounding structures like this one. > > I'd also say that we should only cast it to your _full structure > if we know we have rev 3 of PPTT. Otherwise we should continue manipulating > it as a struct acpi_pptt_cache Fair enough. My thinking was that you had to check the valid flag anyway to use cache_id but it's less robust. I'll delay the casting to later which IIUC is what Jeremy Linton suggested offline. > >> --- >> Changes since v3: >> New patch >> --- >> drivers/acpi/pptt.c | 104 ++++++++++++++++++++++++-------------------- >> 1 file changed, 58 insertions(+), 46 deletions(-) >> >> diff --git a/drivers/acpi/pptt.c b/drivers/acpi/pptt.c >> index 1027ca3566b1..1ed2099c0d1a 100644 >> --- a/drivers/acpi/pptt.c >> +++ b/drivers/acpi/pptt.c >> @@ -21,6 +21,11 @@ >> #include >> #include >> >> +struct acpi_pptt_cache_v1_full { >> + struct acpi_pptt_cache f; >> + struct acpi_pptt_cache_v1 extra; >> +} __packed; > >> +#define ACPI_PPTT_CACHE_V1_LEN sizeof(struct acpi_pptt_cache_v1_full) >> + >> +/* >> + * From PPTT table version 3, a new field cache_id was added at the end of >> + * the cache type structure. We now use struct acpi_pptt_cache_v1_full, >> + * containing the cache_id, everywhere but must check validity before accessing >> + * the cache_id. >> + */ >> +static bool acpi_pptt_cache_id_is_valid(struct acpi_pptt_cache_v1_full *cache) >> +{ >> + return (cache->f.header.length >= ACPI_PPTT_CACHE_V1_LEN && > > Although I later say I don't think you should pass a v1_full structure in here (as > we don't know it is at least that large until after this check) if you do keep this > why not use sizeof(*cache) and get rid of the V1_LEN definition as providing no obvious > value here? Yes, the define was never needed. > >> + cache->f.flags & ACPI_PPTT_CACHE_ID_VALID); >> } > >> @@ -355,7 +374,6 @@ static struct acpi_pptt_cache *acpi_find_cache_node(struct acpi_table_header *ta >> * @this_leaf: Kernel cache info structure being updated >> * @found_cache: The PPTT node describing this cache instance >> * @cpu_node: A unique reference to describe this cache instance >> - * @revision: The revision of the PPTT table >> * >> * The ACPI spec implies that the fields in the cache structures are used to >> * extend and correct the information probed from the hardware. Lets only >> @@ -364,23 +382,20 @@ static struct acpi_pptt_cache *acpi_find_cache_node(struct acpi_table_header *ta >> * Return: nothing. Side effect of updating the global cacheinfo >> */ >> static void update_cache_properties(struct cacheinfo *this_leaf, >> - struct acpi_pptt_cache *found_cache, >> - struct acpi_pptt_processor *cpu_node, >> - u8 revision) >> + struct acpi_pptt_cache_v1_full *found_cache, >> + struct acpi_pptt_processor *cpu_node) >> { >> - struct acpi_pptt_cache_v1* found_cache_v1; >> - >> this_leaf->fw_token = cpu_node; >> - if (found_cache->flags & ACPI_PPTT_SIZE_PROPERTY_VALID) >> - this_leaf->size = found_cache->size; >> - if (found_cache->flags & ACPI_PPTT_LINE_SIZE_VALID) >> - this_leaf->coherency_line_size = found_cache->line_size; >> - if (found_cache->flags & ACPI_PPTT_NUMBER_OF_SETS_VALID) >> - this_leaf->number_of_sets = found_cache->number_of_sets; >> - if (found_cache->flags & ACPI_PPTT_ASSOCIATIVITY_VALID) >> - this_leaf->ways_of_associativity = found_cache->associativity; >> - if (found_cache->flags & ACPI_PPTT_WRITE_POLICY_VALID) { >> - switch (found_cache->attributes & ACPI_PPTT_MASK_WRITE_POLICY) { >> + if (found_cache->f.flags & ACPI_PPTT_SIZE_PROPERTY_VALID) >> + this_leaf->size = found_cache->f.size; >> + if (found_cache->f.flags & ACPI_PPTT_LINE_SIZE_VALID) >> + this_leaf->coherency_line_size = found_cache->f.line_size; >> + if (found_cache->f.flags & ACPI_PPTT_NUMBER_OF_SETS_VALID) >> + this_leaf->number_of_sets = found_cache->f.number_of_sets; >> + if (found_cache->f.flags & ACPI_PPTT_ASSOCIATIVITY_VALID) >> + this_leaf->ways_of_associativity = found_cache->f.associativity; >> + if (found_cache->f.flags & ACPI_PPTT_WRITE_POLICY_VALID) { >> + switch (found_cache->f.attributes & ACPI_PPTT_MASK_WRITE_POLICY) { >> case ACPI_PPTT_CACHE_POLICY_WT: >> this_leaf->attributes = CACHE_WRITE_THROUGH; >> break; >> @@ -389,8 +404,8 @@ static void update_cache_properties(struct cacheinfo *this_leaf, >> break; >> } >> } >> - if (found_cache->flags & ACPI_PPTT_ALLOCATION_TYPE_VALID) { >> - switch (found_cache->attributes & ACPI_PPTT_MASK_ALLOCATION_TYPE) { >> + if (found_cache->f.flags & ACPI_PPTT_ALLOCATION_TYPE_VALID) { >> + switch (found_cache->f.attributes & ACPI_PPTT_MASK_ALLOCATION_TYPE) { >> case ACPI_PPTT_CACHE_READ_ALLOCATE: >> this_leaf->attributes |= CACHE_READ_ALLOCATE; >> break; >> @@ -415,13 +430,11 @@ static void update_cache_properties(struct cacheinfo *this_leaf, >> * specified in PPTT. >> */ >> if (this_leaf->type == CACHE_TYPE_NOCACHE && >> - found_cache->flags & ACPI_PPTT_CACHE_TYPE_VALID) >> + found_cache->f.flags & ACPI_PPTT_CACHE_TYPE_VALID) >> this_leaf->type = CACHE_TYPE_UNIFIED; >> >> - if (revision >= 3 && (found_cache->flags & ACPI_PPTT_CACHE_ID_VALID)) { >> - found_cache_v1 = ACPI_ADD_PTR(struct acpi_pptt_cache_v1, >> - found_cache, sizeof(struct acpi_pptt_cache)); >> - this_leaf->id = found_cache_v1->cache_id; >> + if (acpi_pptt_cache_id_is_valid(found_cache)) { > > Only here do we know that found_cache is the _full type. > >> + this_leaf->id = found_cache->extra.cache_id; >> this_leaf->attributes |= CACHE_ID; >> } >> } >> @@ -429,7 +442,7 @@ static void update_cache_properties(struct cacheinfo *this_leaf, >> static void cache_setup_acpi_cpu(struct acpi_table_header *table, >> unsigned int cpu) >> { >> - struct acpi_pptt_cache *found_cache; >> + struct acpi_pptt_cache_v1_full *found_cache; > > This isn't necessarily valid. Until deep in update_cache_properties() we don't care about the ID > so this structure may be smaller than this implies. > >> struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu); >> u32 acpi_cpu_id = get_acpi_id_for_cpu(cpu); >> struct cacheinfo *this_leaf; >> @@ -445,8 +458,7 @@ static void cache_setup_acpi_cpu(struct acpi_table_header *table, >> pr_debug("found = %p %p\n", found_cache, cpu_node); >> if (found_cache) >> update_cache_properties(this_leaf, found_cache, >> - ACPI_TO_POINTER(ACPI_PTR_DIFF(cpu_node, table)), >> - table->revision); >> + ACPI_TO_POINTER(ACPI_PTR_DIFF(cpu_node, table))); >> >> index++; >> } > Thanks, Ben