From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.remote-ip=91.218.175.95 ARC-Seal: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; cv=none; t=1793971807; b=ViKrGKkwmMVGBfN9Y1dYTrKeZ8yTji4uCH7JZOk2eix5FM3ab41b+iZqZOx652Znee24 RNMenYb+pRfEXcHXGOezqz1rGQhxZxu/B0AatmFyuwFJUh0RGF8WToDNN8kc8+JQhj0pl 7snOfKRGpzLR9XhbM2maUYSp+237iqtx7Z67KGTb1d4uwkGhYIgvx5f6Q2xoe8BAw0BHN fMwuo0JclCOB5n86UyQJ4lN6BMjb2D+07HFeeORW5Jnq+5PX6/ZUFsxidGDT0214WQY1x Qn8vLfK2vvUahWH+On+Trz6VjWXwAxCW9BDXBkw9RNwC+lOPSoJpdR/9d9xglV16RCw== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1793971807; h=DMARC-Filter:X-Envelope-To:DKIM-Signature:X-Envelope-To:Received: X-Mizu-Trace-ID:X-Migadu-Flow:From:To:Cc:Subject:Date:Message-ID: X-Mailer:In-Reply-To:References:MIME-Version: Content-Transfer-Encoding; bh=nChgFP7Zug4DxELreyjciF6D70tMQtD0xbbb+J5QLbE=; b=QpIovzMdgcWyOCZ0fxpCQWg90mr/tiXKoDWFIAHkFj0kW4oI9v/3Szn8PiZQvHYMQ1UA qG85l5ohnRQFXY3hToHzNvfTvDqb3FvElBEl/ZVo7EpYka7FjPAi60MMi0mDVPod2ntgM 5FVClKEFKzXHNFMVPrG8yLNH4mvDGkZhpMBte/0MgtgXvGnfB9WtseL6T3MohUbrh4vuC wAgzeYSmk7Ng8EHbPlC2r/Ykz8n9A56tZCrRIiYZXI8a0wJxZJMtWh8zVjfLgWxjJ2ZsI BC1+nsjxUy0g2B5Ds0y/sh/XTyOtRcxvJGeyoMxK8ZgJ0I3HHSqSiQZZfTAIvhwPmEw== ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass header.d=linux.dev header.i=@linux.dev header.a=rsa-sha256 header.s=key1 header.b=jp0iRFjR; arc=none smtp.remote-ip=91.218.175.95 Received: from mta0.migadu.com (out-95.mta0.migadu.com [91.218.175.95]) (using TLSv1.3 with cipher TLS_AES_128_GCM_SHA256 (128/128 bits) key-exchange X25519 server-signature ECDSA (prime256v1) server-digest SHA256) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id AD9B2606D806 for ; Fri, 6 Nov 2026 13:30:06 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.2 smtp.subspace.kernel.org AD9B2606D806 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key, unprotected) header.d=linux.dev header.i=@linux.dev header.a=rsa-sha256 header.s=key1 header.b=jp0iRFjR X-Envelope-To: linux-kernel@vger.kernel.org DKIM-Signature: a=rsa-sha256; bh=Ok/B0BRR1DxBjJDHlW9S+C61DNS/JMD+2uYdPWi7sl0=; c=simple/simple; d=linux.dev; h=from:to:subject:date:message-id:mime-version:content-type; s=key1; t=1793971804; v=1; x=1794576604; b=jp0iRFjRFlAFv8nlLtO+McWsSiUnsIwJ9gE9BjVvmEkQ/BGyLQZoLP7qh7hs8A25rzNH5T3S MDWvYxcIEuqsZfPhvn2hTFPUq+S9rvqkYb50PWQFHKmAiaMbs8JozEWu/v5fKCaiLSiMwdYGcOF EMhHKTgt5lOi/16Bm00SpVuY= X-Envelope-To: linux-kernel@vger.kernel.org Received: by smtp.migadu.com with ESMTPS id 0e491b9c63b601d3; Fri, 06 Nov 2026 13:30:04 +0000 X-Mizu-Trace-ID: 0e491b9c63b601d3 X-Migadu-Flow: FLOW_OUT From: Fuad Tabba To: Marc Zyngier , Oliver Upton Cc: kvmarm@lists.linux.dev, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, Joey Gouly , Steffen Eiden , Suzuki K Poulose , Zenghui Yu , Catalin Marinas , Will Deacon , Mark Rutland , Raghavendra Rao Ananta , Vincent Donnefort , Wei-Lin Chang , Fuad Tabba Subject: [PATCH v3 1/2] KVM: arm64: Resume stage-2 teardown where the walk stopped Date: Fri, 6 Nov 2026 13:29:59 +0000 Message-ID: <20261106133000.453240-2-fuad.tabba@linux.dev> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20261106133000.453240-1-fuad.tabba@linux.dev> References: <20261106133000.453240-1-fuad.tabba@linux.dev> X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Since commit 4ddfab5436b6 ("KVM: arm64: Reschedule as needed when destroying the stage-2 page-tables"), kvm_stage2_destroy() walks the IPA space one block-sized chunk at a time. The walker steps over an invalid entry in one go, but the next walk starts at the following chunk, so an invalid entry larger than a chunk is visited once for every chunk it covers. Destroying a VM with a 52-bit IPA then takes 2^27 walks with 16K pages (2^22 with 4K), however little it maps. Have kvm_pgtable_stage2_destroy_range() return the address its walk reached, which kvm_pgtable_walk() doesn't expose, and resume from there. A failed walk returns its error instead, and the callers warn about it as the walk used to. stage2_destroy_range() then goes on to the next chunk, as before. pkvm_pgtable_stage2_destroy_range() returns the end of its range. Suggested-by: Suzuki K Poulose Suggested-by: Oliver Upton Signed-off-by: Fuad Tabba --- arch/arm64/include/asm/kvm_pgtable.h | 8 ++++++-- arch/arm64/include/asm/kvm_pkvm.h | 4 ++-- arch/arm64/kvm/hyp/pgtable.c | 19 +++++++++++++++---- arch/arm64/kvm/mmu.c | 6 ++++-- arch/arm64/kvm/pkvm.c | 8 +++++--- 5 files changed, 32 insertions(+), 13 deletions(-) diff --git a/arch/arm64/include/asm/kvm_pgtable.h b/arch/arm64/include/asm/kvm_pgtable.h index 1c9c57cd6ae40..41b137d39a5bb 100644 --- a/arch/arm64/include/asm/kvm_pgtable.h +++ b/arch/arm64/include/asm/kvm_pgtable.h @@ -599,9 +599,13 @@ void kvm_pgtable_stage2_destroy(struct kvm_pgtable *pgt); * * The page-table is assumed to be unreachable by any hardware walkers prior * to freeing and therefore no TLB invalidation is performed. + * + * Return: the address the walk reached, which can be past @addr + @size + * when the range ends within a block or an invalid entry, or a negative + * error code on failure. */ -void kvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, - u64 addr, u64 size); +s64 kvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, + u64 addr, u64 size); /** * kvm_pgtable_stage2_destroy_pgd() - Destroy the PGD of guest stage-2 page-table. diff --git a/arch/arm64/include/asm/kvm_pkvm.h b/arch/arm64/include/asm/kvm_pkvm.h index fd042d2b54624..971ab6c0380de 100644 --- a/arch/arm64/include/asm/kvm_pkvm.h +++ b/arch/arm64/include/asm/kvm_pkvm.h @@ -349,8 +349,8 @@ int __pkvm_handle_smccc_req(struct arm_smccc_res *res); int pkvm_pgtable_stage2_init(struct kvm_pgtable *pgt, struct kvm_s2_mmu *mmu, struct kvm_pgtable_mm_ops *mm_ops); -void pkvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, - u64 addr, u64 size); +s64 pkvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, + u64 addr, u64 size); void pkvm_pgtable_stage2_destroy_pgd(struct kvm_pgtable *pgt); int pkvm_pgtable_stage2_map(struct kvm_pgtable *pgt, u64 addr, u64 size, u64 phys, enum kvm_pgtable_prot prot, void *mc, diff --git a/arch/arm64/kvm/hyp/pgtable.c b/arch/arm64/kvm/hyp/pgtable.c index 9c83837b184d3..01fc3fc2b6553 100644 --- a/arch/arm64/kvm/hyp/pgtable.c +++ b/arch/arm64/kvm/hyp/pgtable.c @@ -1653,16 +1653,27 @@ static int stage2_free_walker(const struct kvm_pgtable_visit_ctx *ctx, } } -void kvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, - u64 addr, u64 size) +s64 kvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, + u64 addr, u64 size) { struct kvm_pgtable_walker walker = { .cb = stage2_free_walker, .flags = KVM_PGTABLE_WALK_LEAF | KVM_PGTABLE_WALK_TABLE_POST, }; + struct kvm_pgtable_walk_data data = { + .walker = &walker, + .start = addr, + .addr = addr, + .end = addr + size, + }; + int ret; - WARN_ON(kvm_pgtable_walk(pgt, addr, size, &walker)); + ret = _kvm_pgtable_walk(pgt, &data); + if (ret) + return ret; + + return data.addr; } void kvm_pgtable_stage2_destroy_pgd(struct kvm_pgtable *pgt) @@ -1681,7 +1692,7 @@ void kvm_pgtable_stage2_destroy_pgd(struct kvm_pgtable *pgt) void kvm_pgtable_stage2_destroy(struct kvm_pgtable *pgt) { - kvm_pgtable_stage2_destroy_range(pgt, 0, BIT(pgt->ia_bits)); + WARN_ON(kvm_pgtable_stage2_destroy_range(pgt, 0, BIT(pgt->ia_bits)) < 0); kvm_pgtable_stage2_destroy_pgd(pgt); } diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c index 1db10d8d6db73..372a1c04d3c47 100644 --- a/arch/arm64/kvm/mmu.c +++ b/arch/arm64/kvm/mmu.c @@ -1008,11 +1008,13 @@ static void stage2_destroy_range(struct kvm_pgtable *pgt, phys_addr_t addr, phys_addr_t end) { u64 next; + s64 ret; do { next = stage2_range_addr_end(addr, end); - KVM_PGT_FN(kvm_pgtable_stage2_destroy_range)(pgt, addr, - next - addr); + ret = KVM_PGT_FN(kvm_pgtable_stage2_destroy_range)(pgt, addr, next - addr); + if (!WARN_ON(ret < 0)) + next = min_t(u64, ret, end); if (next != end) cond_resched(); } while (addr = next, addr != end); diff --git a/arch/arm64/kvm/pkvm.c b/arch/arm64/kvm/pkvm.c index 016997288fd23..d0a6308008512 100644 --- a/arch/arm64/kvm/pkvm.c +++ b/arch/arm64/kvm/pkvm.c @@ -499,14 +499,14 @@ static int __pkvm_pgtable_stage2_unshare(struct kvm_pgtable *pgt, u64 start, u64 return 0; } -void pkvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, - u64 addr, u64 size) +s64 pkvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, + u64 addr, u64 size) { struct kvm *kvm = kvm_s2_mmu_to_kvm(pgt->mmu); pkvm_handle_t handle = kvm->arch.pkvm.handle; if (!handle) - return; + return addr + size; if (pkvm_hyp_vm_is_created(kvm) && !kvm->arch.pkvm.is_dying) { WARN_ON(kvm_call_hyp_nvhe(__pkvm_start_teardown_vm, handle)); @@ -517,6 +517,8 @@ void pkvm_pgtable_stage2_destroy_range(struct kvm_pgtable *pgt, __pkvm_pgtable_stage2_reclaim(pgt, addr, addr + size); else __pkvm_pgtable_stage2_unshare(pgt, addr, addr + size); + + return addr + size; } void pkvm_pgtable_stage2_destroy_pgd(struct kvm_pgtable *pgt) -- 2.47.3