From nobody Fri Dec 19 04:03:16 2025 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 ACA91CA0EC3 for ; Mon, 11 Sep 2023 21:10:23 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231842AbjIKVE5 (ORCPT ); Mon, 11 Sep 2023 17:04:57 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:44400 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237858AbjIKNPi (ORCPT ); Mon, 11 Sep 2023 09:15:38 -0400 Received: from mail-wm1-x330.google.com (mail-wm1-x330.google.com [IPv6:2a00:1450:4864:20::330]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 93519EB for ; Mon, 11 Sep 2023 06:15:32 -0700 (PDT) Received: by mail-wm1-x330.google.com with SMTP id 5b1f17b1804b1-403012f27e1so23418325e9.1 for ; Mon, 11 Sep 2023 06:15:32 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=rivosinc-com.20230601.gappssmtp.com; s=20230601; t=1694438131; x=1695042931; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=nqYb5Vu8dCDUBKbfeLnWwY3UtdJspJqfkbvYUWs56pc=; b=E+tslBdIxbMHSUp30W8BlVy29pmiEBopVhkQQs02cim9x5GoOj9pnrnTyhPKOaaZEZ +kjgqOscEvCVTTuwX2EnyNSDpZQzTCE+/mPQf0HAmWXz2d388r20YL+STC16B0gR/jhI ax7W52zskGl8UsLtmJcE/XtfNS09yWhQscmL/0tKLbTXOx2i7/Wd8NpLelVyAnmzcuWR 25xDClpNjGEaZ4QY3LbyqEKQ0hqCDDDVEYoDQZeGXFylUFjnGfHPV+F6chpCwwp6RaG+ u7WP7Yw4gmXXyrGby6xM3wzqU+S9LBq89hBbZUgyzQnJTFoDjDCnW/Fsundp9ZfJQFRn seVg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1694438131; x=1695042931; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=nqYb5Vu8dCDUBKbfeLnWwY3UtdJspJqfkbvYUWs56pc=; b=c65+FLHY1BNmitaNIhQZF0rg1MeWjaNzYMY7FjxvYm01XocIoRNcf5Qcr6JomH1Lpl 31Wo9xM+M7+jaO8mjv0mmrC7OuGax9hQ6OxMJXroqsVyodZKNJMQDFz/jCo8XTKuY/1G yLYjxKmPLoJTMAL/Imd28/qBB2NN9s0OM8zMl3CJ+bDkSPF4JozE9xBg7pCEO0k4sc/x BcVGMT3CS8ZaF6JNI5+zlSNbJ+Y4WQPMtw1bPOEArW0LxkK0gZo7bzGUJjtIOJnqimow HTOlubPbG9/UcFBsWrc4ddLi41445/YQ33XJkW4v+bjYH+JxZJoYdJOSGcqxuZTiJJMO vXWA== X-Gm-Message-State: AOJu0YwR8Bext6M0iSX6lY4fxNSrn0C4kgHke7SEVgNNdqlMBgnXQGb3 BtMeoxv9LKhXYhilox8vdcjkZA== X-Google-Smtp-Source: AGHT+IGSM7mk4P/JNlSdasDtYzZnQXEd3SbuOn7pwwi1IoVXpt21uzh3Nx2y9SMVd5/QCh9kvlN/TA== X-Received: by 2002:a05:600c:2301:b0:400:ce4f:f184 with SMTP id 1-20020a05600c230100b00400ce4ff184mr8883368wmo.41.1694438130933; Mon, 11 Sep 2023 06:15:30 -0700 (PDT) Received: from alex-rivos.ba.rivosinc.com (amontpellier-656-1-456-62.w92-145.abo.wanadoo.fr. [92.145.124.62]) by smtp.gmail.com with ESMTPSA id m31-20020a05600c3b1f00b00402f7e473b7sm9878654wms.15.2023.09.11.06.15.30 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 11 Sep 2023 06:15:30 -0700 (PDT) From: Alexandre Ghiti To: Will Deacon , "Aneesh Kumar K . V" , Andrew Morton , Nick Piggin , Peter Zijlstra , Mayuresh Chitale , Vincent Chen , Paul Walmsley , Palmer Dabbelt , Albert Ou , linux-arch@vger.kernel.org, linux-mm@kvack.org, linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, Samuel Holland , Lad Prabhakar Cc: Alexandre Ghiti , Andrew Jones Subject: [PATCH v4 3/4] riscv: Make __flush_tlb_range() loop over pte instead of flushing the whole tlb Date: Mon, 11 Sep 2023 15:12:23 +0200 Message-Id: <20230911131224.61924-4-alexghiti@rivosinc.com> X-Mailer: git-send-email 2.39.2 In-Reply-To: <20230911131224.61924-1-alexghiti@rivosinc.com> References: <20230911131224.61924-1-alexghiti@rivosinc.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Currently, when the range to flush covers more than one page (a 4K page or a hugepage), __flush_tlb_range() flushes the whole tlb. Flushing the whole tlb comes with a greater cost than flushing a single entry so we should flush single entries up to a certain threshold so that: threshold * cost of flushing a single entry < cost of flushing the whole tlb. Co-developed-by: Mayuresh Chitale Signed-off-by: Mayuresh Chitale Signed-off-by: Alexandre Ghiti Reviewed-by: Andrew Jones Tested-by: Lad Prabhakar # --- arch/riscv/include/asm/sbi.h | 3 - arch/riscv/include/asm/tlbflush.h | 3 + arch/riscv/kernel/sbi.c | 32 +++------ arch/riscv/mm/tlbflush.c | 115 +++++++++++++++--------------- 4 files changed, 72 insertions(+), 81 deletions(-) diff --git a/arch/riscv/include/asm/sbi.h b/arch/riscv/include/asm/sbi.h index 5b4a1bf5f439..b79d0228144f 100644 --- a/arch/riscv/include/asm/sbi.h +++ b/arch/riscv/include/asm/sbi.h @@ -273,9 +273,6 @@ void sbi_set_timer(uint64_t stime_value); void sbi_shutdown(void); void sbi_send_ipi(unsigned int cpu); int sbi_remote_fence_i(const struct cpumask *cpu_mask); -int sbi_remote_sfence_vma(const struct cpumask *cpu_mask, - unsigned long start, - unsigned long size); =20 int sbi_remote_sfence_vma_asid(const struct cpumask *cpu_mask, unsigned long start, diff --git a/arch/riscv/include/asm/tlbflush.h b/arch/riscv/include/asm/tlb= flush.h index f5c4fb0ae642..170a49c531c6 100644 --- a/arch/riscv/include/asm/tlbflush.h +++ b/arch/riscv/include/asm/tlbflush.h @@ -11,6 +11,9 @@ #include #include =20 +#define FLUSH_TLB_MAX_SIZE ((unsigned long)-1) +#define FLUSH_TLB_NO_ASID ((unsigned long)-1) + #ifdef CONFIG_MMU extern unsigned long asid_mask; =20 diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c index c672c8ba9a2a..5a62ed1da453 100644 --- a/arch/riscv/kernel/sbi.c +++ b/arch/riscv/kernel/sbi.c @@ -11,6 +11,7 @@ #include #include #include +#include =20 /* default SBI version is 0.1 */ unsigned long sbi_spec_version __ro_after_init =3D SBI_SPEC_VERSION_DEFAUL= T; @@ -376,32 +377,15 @@ int sbi_remote_fence_i(const struct cpumask *cpu_mask) } EXPORT_SYMBOL(sbi_remote_fence_i); =20 -/** - * sbi_remote_sfence_vma() - Execute SFENCE.VMA instructions on given remo= te - * harts for the specified virtual address range. - * @cpu_mask: A cpu mask containing all the target harts. - * @start: Start of the virtual address - * @size: Total size of the virtual address range. - * - * Return: 0 on success, appropriate linux error code otherwise. - */ -int sbi_remote_sfence_vma(const struct cpumask *cpu_mask, - unsigned long start, - unsigned long size) -{ - return __sbi_rfence(SBI_EXT_RFENCE_REMOTE_SFENCE_VMA, - cpu_mask, start, size, 0, 0); -} -EXPORT_SYMBOL(sbi_remote_sfence_vma); - /** * sbi_remote_sfence_vma_asid() - Execute SFENCE.VMA instructions on given - * remote harts for a virtual address range belonging to a specific ASID. + * remote harts for a virtual address range belonging to a specific ASID o= r not. * * @cpu_mask: A cpu mask containing all the target harts. * @start: Start of the virtual address * @size: Total size of the virtual address range. - * @asid: The value of address space identifier (ASID). + * @asid: The value of address space identifier (ASID), or FLUSH_TLB_NO_AS= ID + * for flushing all address spaces. * * Return: 0 on success, appropriate linux error code otherwise. */ @@ -410,8 +394,12 @@ int sbi_remote_sfence_vma_asid(const struct cpumask *c= pu_mask, unsigned long size, unsigned long asid) { - return __sbi_rfence(SBI_EXT_RFENCE_REMOTE_SFENCE_VMA_ASID, - cpu_mask, start, size, asid, 0); + if (asid =3D=3D FLUSH_TLB_NO_ASID) + return __sbi_rfence(SBI_EXT_RFENCE_REMOTE_SFENCE_VMA, + cpu_mask, start, size, 0, 0); + else + return __sbi_rfence(SBI_EXT_RFENCE_REMOTE_SFENCE_VMA_ASID, + cpu_mask, start, size, asid, 0); } EXPORT_SYMBOL(sbi_remote_sfence_vma_asid); =20 diff --git a/arch/riscv/mm/tlbflush.c b/arch/riscv/mm/tlbflush.c index 5bda6d4fed90..2c1136d73411 100644 --- a/arch/riscv/mm/tlbflush.c +++ b/arch/riscv/mm/tlbflush.c @@ -9,28 +9,50 @@ =20 static inline void local_flush_tlb_all_asid(unsigned long asid) { - __asm__ __volatile__ ("sfence.vma x0, %0" - : - : "r" (asid) - : "memory"); + if (asid !=3D FLUSH_TLB_NO_ASID) + __asm__ __volatile__ ("sfence.vma x0, %0" + : + : "r" (asid) + : "memory"); + else + local_flush_tlb_all(); } =20 static inline void local_flush_tlb_page_asid(unsigned long addr, unsigned long asid) { - __asm__ __volatile__ ("sfence.vma %0, %1" - : - : "r" (addr), "r" (asid) - : "memory"); + if (asid !=3D FLUSH_TLB_NO_ASID) + __asm__ __volatile__ ("sfence.vma %0, %1" + : + : "r" (addr), "r" (asid) + : "memory"); + else + local_flush_tlb_page(addr); } =20 -static inline void local_flush_tlb_range(unsigned long start, - unsigned long size, unsigned long stride) +/* + * Flush entire TLB if number of entries to be flushed is greater + * than the threshold below. + */ +static unsigned long tlb_flush_all_threshold __read_mostly =3D 64; + +static void local_flush_tlb_range_threshold_asid(unsigned long start, + unsigned long size, + unsigned long stride, + unsigned long asid) { - if (size <=3D stride) - local_flush_tlb_page(start); - else - local_flush_tlb_all(); + u16 nr_ptes_in_range =3D DIV_ROUND_UP(size, stride); + int i; + + if (nr_ptes_in_range > tlb_flush_all_threshold) { + local_flush_tlb_all_asid(asid); + return; + } + + for (i =3D 0; i < nr_ptes_in_range; ++i) { + local_flush_tlb_page_asid(start, asid); + start +=3D stride; + } } =20 static inline void local_flush_tlb_range_asid(unsigned long start, @@ -38,8 +60,10 @@ static inline void local_flush_tlb_range_asid(unsigned l= ong start, { if (size <=3D stride) local_flush_tlb_page_asid(start, asid); - else + else if (size =3D=3D FLUSH_TLB_MAX_SIZE) local_flush_tlb_all_asid(asid); + else + local_flush_tlb_range_threshold_asid(start, size, stride, asid); } =20 static void __ipi_flush_tlb_all(void *info) @@ -52,7 +76,7 @@ void flush_tlb_all(void) if (riscv_use_ipi_for_rfence()) on_each_cpu(__ipi_flush_tlb_all, NULL, 1); else - sbi_remote_sfence_vma(NULL, 0, -1); + sbi_remote_sfence_vma_asid(NULL, 0, FLUSH_TLB_MAX_SIZE, FLUSH_TLB_NO_ASI= D); } =20 struct flush_tlb_range_data { @@ -69,18 +93,12 @@ static void __ipi_flush_tlb_range_asid(void *info) local_flush_tlb_range_asid(d->start, d->size, d->stride, d->asid); } =20 -static void __ipi_flush_tlb_range(void *info) -{ - struct flush_tlb_range_data *d =3D info; - - local_flush_tlb_range(d->start, d->size, d->stride); -} - static void __flush_tlb_range(struct mm_struct *mm, unsigned long start, unsigned long size, unsigned long stride) { struct flush_tlb_range_data ftd; struct cpumask *cmask =3D mm_cpumask(mm); + unsigned long asid =3D FLUSH_TLB_NO_ASID; unsigned int cpuid; bool broadcast; =20 @@ -90,39 +108,24 @@ static void __flush_tlb_range(struct mm_struct *mm, un= signed long start, cpuid =3D get_cpu(); /* check if the tlbflush needs to be sent to other CPUs */ broadcast =3D cpumask_any_but(cmask, cpuid) < nr_cpu_ids; - if (static_branch_unlikely(&use_asid_allocator)) { - unsigned long asid =3D atomic_long_read(&mm->context.id) & asid_mask; - - if (broadcast) { - if (riscv_use_ipi_for_rfence()) { - ftd.asid =3D asid; - ftd.start =3D start; - ftd.size =3D size; - ftd.stride =3D stride; - on_each_cpu_mask(cmask, - __ipi_flush_tlb_range_asid, - &ftd, 1); - } else - sbi_remote_sfence_vma_asid(cmask, - start, size, asid); - } else { - local_flush_tlb_range_asid(start, size, stride, asid); - } + + if (static_branch_unlikely(&use_asid_allocator)) + asid =3D atomic_long_read(&mm->context.id) & asid_mask; + + if (broadcast) { + if (riscv_use_ipi_for_rfence()) { + ftd.asid =3D asid; + ftd.start =3D start; + ftd.size =3D size; + ftd.stride =3D stride; + on_each_cpu_mask(cmask, + __ipi_flush_tlb_range_asid, + &ftd, 1); + } else + sbi_remote_sfence_vma_asid(cmask, + start, size, asid); } else { - if (broadcast) { - if (riscv_use_ipi_for_rfence()) { - ftd.asid =3D 0; - ftd.start =3D start; - ftd.size =3D size; - ftd.stride =3D stride; - on_each_cpu_mask(cmask, - __ipi_flush_tlb_range, - &ftd, 1); - } else - sbi_remote_sfence_vma(cmask, start, size); - } else { - local_flush_tlb_range(start, size, stride); - } + local_flush_tlb_range_asid(start, size, stride, asid); } =20 put_cpu(); @@ -130,7 +133,7 @@ static void __flush_tlb_range(struct mm_struct *mm, uns= igned long start, =20 void flush_tlb_mm(struct mm_struct *mm) { - __flush_tlb_range(mm, 0, -1, PAGE_SIZE); + __flush_tlb_range(mm, 0, FLUSH_TLB_MAX_SIZE, PAGE_SIZE); } =20 void flush_tlb_mm_range(struct mm_struct *mm, --=20 2.39.2