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 81A3FC001DE for ; Tue, 25 Jul 2023 12:03:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S233237AbjGYMDu (ORCPT ); Tue, 25 Jul 2023 08:03:50 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40578 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233089AbjGYMDp (ORCPT ); Tue, 25 Jul 2023 08:03:45 -0400 Received: from szxga01-in.huawei.com (szxga01-in.huawei.com [45.249.212.187]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3517CA3 for ; Tue, 25 Jul 2023 05:03:43 -0700 (PDT) Received: from canpemm500005.china.huawei.com (unknown [172.30.72.57]) by szxga01-in.huawei.com (SkyGuard) with ESMTP id 4R9Fyw5dyDzrRtD; Tue, 25 Jul 2023 20:02:48 +0800 (CST) Received: from [10.67.110.73] (10.67.110.73) by canpemm500005.china.huawei.com (7.192.104.229) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Tue, 25 Jul 2023 20:03:40 +0800 Message-ID: <32552071-9e0f-b86b-202a-5cdf20871f59@huawei.com> Date: Tue, 25 Jul 2023 20:03:40 +0800 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101 Thunderbird/102.3.3 Subject: Re: [PATCH -next v2] arm64: fix build warning for ARM64_MEMSTART_SHIFT To: Catalin Marinas CC: Anshuman Khandual , , , , , , , References: <20230724172751.3048501-1-chris.zjh@huawei.com> <4938685e-d293-4970-b964-3b9f560d45dd@arm.com> <9fcfe47f-9289-8eb5-ce4e-9f66648b0e89@huawei.com> From: "zhangjianhua (E)" In-Reply-To: Content-Type: text/plain; charset="UTF-8"; format=flowed Content-Transfer-Encoding: 8bit X-Originating-IP: [10.67.110.73] X-ClientProxiedBy: dggems703-chm.china.huawei.com (10.3.19.180) To canpemm500005.china.huawei.com (7.192.104.229) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 在 2023/7/25 18:15, Catalin Marinas 写道: > On Tue, Jul 25, 2023 at 04:47:46PM +0800, zhangjianhua (E) wrote: >> 在 2023/7/25 12:22, Anshuman Khandual 写道: >>> On 7/24/23 20:41, Catalin Marinas wrote: >>>> On Mon, Jul 24, 2023 at 05:27:51PM +0000, Zhang Jianhua wrote: >>>>> When building with W=1, the following warning occurs. >>>>> >>>>> arch/arm64/include/asm/kernel-pgtable.h:129:41: error: "PUD_SHIFT" is not defined, evaluates to 0 [-Werror=undef] >>>>> 129 | #define ARM64_MEMSTART_SHIFT PUD_SHIFT >>>>> | ^~~~~~~~~ >>>>> arch/arm64/include/asm/kernel-pgtable.h:142:5: note: in expansion of macro ‘ARM64_MEMSTART_SHIFT’ >>>>> 142 | #if ARM64_MEMSTART_SHIFT < SECTION_SIZE_BITS >>>>> | ^~~~~~~~~~~~~~~~~~~~ >>>> Another thing that's missing here is that the warning is probably when >>>> this file is included from asm-offests.h or some .S file. >>>> >>>>> diff --git a/arch/arm64/include/asm/kernel-pgtable.h b/arch/arm64/include/asm/kernel-pgtable.h >>>>> index 577773870b66..51bdce66885d 100644 >>>>> --- a/arch/arm64/include/asm/kernel-pgtable.h >>>>> +++ b/arch/arm64/include/asm/kernel-pgtable.h >>>>> @@ -125,12 +125,14 @@ >>>>> * (64k granule), or a multiple that can be mapped using contiguous bits >>>>> * in the page tables: 32 * PMD_SIZE (16k granule) >>>>> */ >>>>> -#if defined(CONFIG_ARM64_4K_PAGES) >>>>> +#if defined(CONFIG_ARM64_4K_PAGES) && defined(PUD_SHIFT) >>>>> #define ARM64_MEMSTART_SHIFT PUD_SHIFT >>>> That's not the correct fix since PUD_SHIFT should always be defined. >>>> When CONFIG_PGTABLE_LEVELS == 3, pgtable-types.h includes >>>> asm-generic/pgtable-nopud.h and this defines PUD_SHIFT. We either got >>> Right, PUD_SHIFT is always defined irrespective of page table levels. >>> >>>> ARM64_MEMSTART_SHIFT defined in the wrong file or kernel-pgtable.h does >>>> not pull the relevant headers (either directly or via an included >>>> header). Even if kernel-pgtable.h ends up including the nopud/nopmd >>>> headers, P*D_SHIFT is guarded by an #indef __ASSEMBLY__ in those files. >>>> >>>> Something like below appears to fix this, though I'm not particularly >>>> fond of guarding the ARM64_MEMSTART_* definitions by #ifndef >>>> __ASSEMBLY__ for no apparent reason (could add a comment though): >>>> >>>> -----------------------8<--------------------------- >>>> diff --git a/arch/arm64/include/asm/kernel-pgtable.h b/arch/arm64/include/asm/kernel-pgtable.h >>>> index 577773870b66..fcea7e87a6ca 100644 >>>> --- a/arch/arm64/include/asm/kernel-pgtable.h >>>> +++ b/arch/arm64/include/asm/kernel-pgtable.h >>>> @@ -118,6 +118,8 @@ >>>> #define SWAPPER_RX_MMUFLAGS (SWAPPER_RW_MMUFLAGS | PTE_RDONLY) >>>> #endif >>>> +#ifndef __ASSEMBLY__ >>>> + >>>> /* >>>> * To make optimal use of block mappings when laying out the linear >>>> * mapping, round down the base of physical memory to a size that can >>>> @@ -145,4 +147,6 @@ >>>> #define ARM64_MEMSTART_ALIGN (1UL << ARM64_MEMSTART_SHIFT) >>>> #endif >>>> +#endif /* __ASSEMBLY__ */ >>>> + >>>> #endif /* __ASM_KERNEL_PGTABLE_H */ >>>> diff --git a/arch/arm64/include/asm/pgtable-hwdef.h b/arch/arm64/include/asm/pgtable-hwdef.h >>>> index e4944d517c99..22b36f2d5d93 100644 >>>> --- a/arch/arm64/include/asm/pgtable-hwdef.h >>>> +++ b/arch/arm64/include/asm/pgtable-hwdef.h >>>> @@ -6,6 +6,7 @@ >>>> #define __ASM_PGTABLE_HWDEF_H >>>> #include >>>> +#include >>>> /* >>>> * Number of page-table levels required to address 'va_bits' wide >>>> diff --git a/arch/arm64/include/asm/pgtable-types.h b/arch/arm64/include/asm/pgtable-types.h >>>> index b8f158ae2527..ae86e66fdb11 100644 >>>> --- a/arch/arm64/include/asm/pgtable-types.h >>>> +++ b/arch/arm64/include/asm/pgtable-types.h >>>> @@ -11,6 +11,8 @@ >>>> #include >>>> +#ifndef __ASSEMBLY__ >>>> + >>>> typedef u64 pteval_t; >>>> typedef u64 pmdval_t; >>>> typedef u64 pudval_t; >>>> @@ -44,6 +46,8 @@ typedef struct { pteval_t pgprot; } pgprot_t; >>>> #define pgprot_val(x) ((x).pgprot) >>>> #define __pgprot(x) ((pgprot_t) { (x) } ) >>>> +#endif /* __ASSEMBLY__ */ >>>> + >>>> #if CONFIG_PGTABLE_LEVELS == 2 >>>> #include >>>> #elif CONFIG_PGTABLE_LEVELS == 3 >>>> -----------------------8<--------------------------- >>>> >>>> To avoid guarding the ARM64_MEMSTART_* definitions, we could instead >>>> move the P*D_SHIFT definitions in asm-generic/pgtable-nop*d.h outside >>>> the #ifndef __ASSEMBLY__ block. >>> OR could ARM64_MEMSTART_SHIFT and ARM64_MEMSTART_ALIGN computation blocks >>> be moved inside arch/arm64/mm/init.c, where it is used exclusively. Seems >>> to be solving the problem as well. > That's fine by me, better than adding the #ifndef __ASSEMBLY__ around > them. > >> This method can avoid the current compilation warning, but does not >> solve the problem that PUD_SHIFT and PMD_SHIFT undefined in fact, and >> it is contrary to XXX_SHIFT should always be defined. Maybe it would >> be more appropriate to solve this issue directly. > For .c files, we can solve this by including asm/pgtable-types.h in > asm/pgtable-hwdef.h. This still leaves P*D_SHIFT undefined for .S files > since the generic nop*d.h headers guard the shifts with !__ASSEMBLY__ > but do we really care about this? I haven't seen any other warning of > P*D_SHIFT not being defined. If you do want to solve these, just go and > change the generic headers to take the shift out of the asm guard. I > don't think it's worth it. OK, agree, I will send a new patch for Anshuman's method.