From nobody Sun Feb 8 17:37:30 2026 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 7EAEFC04E69 for ; Tue, 8 Aug 2023 23:14:38 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231805AbjHHXOh (ORCPT ); Tue, 8 Aug 2023 19:14:37 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:49666 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232137AbjHHXOS (ORCPT ); Tue, 8 Aug 2023 19:14:18 -0400 Received: from mail-yb1-xb49.google.com (mail-yb1-xb49.google.com [IPv6:2607:f8b0:4864:20::b49]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 53D4E212D for ; Tue, 8 Aug 2023 16:13:50 -0700 (PDT) Received: by mail-yb1-xb49.google.com with SMTP id 3f1490d57ef6-d60d944f40bso173730276.3 for ; Tue, 08 Aug 2023 16:13:50 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20221208; t=1691536429; x=1692141229; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=tRhPhhyCVrvwh8HFH9MoQf2shI38NIr8PShGiYFRBxo=; b=Zc3XMNBSUBa7GJhE2tyVMoUFP74xH5x2uC+RObeuMypOT4zBfhNjgfLsgmCF/VIUL/ c3YFkKJLvbJ6noh+FEA5EU6B4Y8bvu62qUzHsuhleSUfz8cH5tQ1h2wfzfwa++wsDdyt FdRoKf2t/zrH8Qz4fT0OA2Kl3es/jcoD04x+E+eEDzWE42d8etHaIXAOrwOpk5u/mKfh SD5EM0YbNXhL7Ht7OKRuV/ykqM5iiJ9AioVsAY1rEO1cFvrWt6QecQuf0sR7bTSV6B5Y 5rkKiGP8xtgnes8eLdl7Hzw4R6tX27rsB7R7EcqhdtJaraNhkYwbavnno/QAT62FPCbn dJBw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20221208; t=1691536429; x=1692141229; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=tRhPhhyCVrvwh8HFH9MoQf2shI38NIr8PShGiYFRBxo=; b=f8GYbtINnAQDOCsyznxVBZpYIyyGhyo1ojd353TnagXP64kWaUoMI3GWC+xXL4a1lg j/w3WifkMYWDYiL+riEkjvDLxS6FSdhz72AryWXCJtkrHhGQnH5AWRBlglog7JaAcyNY kDdYShWX/G2LDP1u7WYCbpH+8YYkqvVVcZFq8X9YywP0dHuXadR4JOPaW7NEXSOGrLXQ O173s6FL2I8Xj9fFuxhWtjBDujUz5K21DbSR5L26DIUxIzXoIhkDziEODcLoSMcm3eBZ 9vV0fcHfbUyI/fTGqaxJEvO56CxEJ7tmsvCuFuGG8PY6HzpW0r8bQSCHhKI8Vlx9Hjl9 jtEA== X-Gm-Message-State: AOJu0YxPIJpr4nIk4WbfTkvczGDv/zYNx3A1VVzyTsCAPQk9U+Q1/uoN ZGTeW88ntvXn2cJ8HRnxwMiCKr3qmwki X-Google-Smtp-Source: AGHT+IEDJhl5pfGMsCT/iI4kA7cKyLhi3xdcFmiU1wBNjxnpzamybXY7e3K9E1XS3MNVQtxNcdIYmyqlwdlH X-Received: from rananta-linux.c.googlers.com ([fda3:e722:ac3:cc00:2b:ff92:c0a8:22b5]) (user=rananta job=sendgmr) by 2002:a5b:88c:0:b0:d1e:721b:469d with SMTP id e12-20020a5b088c000000b00d1e721b469dmr20600ybq.7.1691536429525; Tue, 08 Aug 2023 16:13:49 -0700 (PDT) Date: Tue, 8 Aug 2023 23:13:30 +0000 In-Reply-To: <20230808231330.3855936-1-rananta@google.com> Mime-Version: 1.0 References: <20230808231330.3855936-1-rananta@google.com> X-Mailer: git-send-email 2.41.0.640.ga95def55d0-goog Message-ID: <20230808231330.3855936-15-rananta@google.com> Subject: [PATCH v8 14/14] KVM: arm64: Use TLBI range-based intructions for unmap From: Raghavendra Rao Ananta To: Oliver Upton , Marc Zyngier , James Morse , Suzuki K Poulose Cc: Paolo Bonzini , Sean Christopherson , Huacai Chen , Zenghui Yu , Anup Patel , Atish Patra , Jing Zhang , Reiji Watanabe , Colton Lewis , Raghavendra Rao Anata , David Matlack , Fuad Tabba , linux-arm-kernel@lists.infradead.org, kvmarm@lists.linux.dev, linux-mips@vger.kernel.org, kvm-riscv@lists.infradead.org, linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, kvm@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" The current implementation of the stage-2 unmap walker traverses the given range and, as a part of break-before-make, performs TLB invalidations with a DSB for every PTE. A multitude of this combination could cause a performance bottleneck on some systems. Hence, if the system supports FEAT_TLBIRANGE, defer the TLB invalidations until the entire walk is finished, and then use range-based instructions to invalidate the TLBs in one go. Condition deferred TLB invalidation on the system supporting FWB, as the optimization is entirely pointless when the unmap walker needs to perform CMOs. Rename stage2_put_pte() to stage2_unmap_put_pte() as the function now serves the stage-2 unmap walker specifically, rather than acting generic. Signed-off-by: Raghavendra Rao Ananta Reviewed-by: Shaoqin Huang --- arch/arm64/kvm/hyp/pgtable.c | 40 +++++++++++++++++++++++++++++------- 1 file changed, 33 insertions(+), 7 deletions(-) diff --git a/arch/arm64/kvm/hyp/pgtable.c b/arch/arm64/kvm/hyp/pgtable.c index 5ef098af17362..eaaae76481fa9 100644 --- a/arch/arm64/kvm/hyp/pgtable.c +++ b/arch/arm64/kvm/hyp/pgtable.c @@ -831,16 +831,36 @@ static void stage2_make_pte(const struct kvm_pgtable_= visit_ctx *ctx, kvm_pte_t n smp_store_release(ctx->ptep, new); } =20 -static void stage2_put_pte(const struct kvm_pgtable_visit_ctx *ctx, struct= kvm_s2_mmu *mmu, - struct kvm_pgtable_mm_ops *mm_ops) +static bool stage2_unmap_defer_tlb_flush(struct kvm_pgtable *pgt) { /* - * Clear the existing PTE, and perform break-before-make with - * TLB maintenance if it was valid. + * If FEAT_TLBIRANGE is implemented, defer the individual + * TLB invalidations until the entire walk is finished, and + * then use the range-based TLBI instructions to do the + * invalidations. Condition deferred TLB invalidation on the + * system supporting FWB as the optimization is entirely + * pointless when the unmap walker needs to perform CMOs. + */ + return system_supports_tlb_range() && stage2_has_fwb(pgt); +} + +static void stage2_unmap_put_pte(const struct kvm_pgtable_visit_ctx *ctx, + struct kvm_s2_mmu *mmu, + struct kvm_pgtable_mm_ops *mm_ops) +{ + struct kvm_pgtable *pgt =3D ctx->arg; + + /* + * Clear the existing PTE, and perform break-before-make if it was + * valid. Depending on the system support, defer the TLB maintenance + * for the same until the entire unmap walk is completed. */ if (kvm_pte_valid(ctx->old)) { kvm_clear_pte(ctx->ptep); - kvm_call_hyp(__kvm_tlb_flush_vmid_ipa, mmu, ctx->addr, ctx->level); + + if (!stage2_unmap_defer_tlb_flush(pgt)) + kvm_call_hyp(__kvm_tlb_flush_vmid_ipa, mmu, + ctx->addr, ctx->level); } =20 mm_ops->put_page(ctx->ptep); @@ -1098,7 +1118,7 @@ static int stage2_unmap_walker(const struct kvm_pgtab= le_visit_ctx *ctx, * block entry and rely on the remaining portions being faulted * back lazily. */ - stage2_put_pte(ctx, mmu, mm_ops); + stage2_unmap_put_pte(ctx, mmu, mm_ops); =20 if (need_flush && mm_ops->dcache_clean_inval_poc) mm_ops->dcache_clean_inval_poc(kvm_pte_follow(ctx->old, mm_ops), @@ -1112,13 +1132,19 @@ static int stage2_unmap_walker(const struct kvm_pgt= able_visit_ctx *ctx, =20 int kvm_pgtable_stage2_unmap(struct kvm_pgtable *pgt, u64 addr, u64 size) { + int ret; struct kvm_pgtable_walker walker =3D { .cb =3D stage2_unmap_walker, .arg =3D pgt, .flags =3D KVM_PGTABLE_WALK_LEAF | KVM_PGTABLE_WALK_TABLE_POST, }; =20 - return kvm_pgtable_walk(pgt, addr, size, &walker); + ret =3D kvm_pgtable_walk(pgt, addr, size, &walker); + if (stage2_unmap_defer_tlb_flush(pgt)) + /* Perform the deferred TLB invalidations */ + kvm_tlb_flush_vmid_range(pgt->mmu, addr, size); + + return ret; } =20 struct stage2_attr_data { --=20 2.41.0.640.ga95def55d0-goog