linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: cdall@linaro.org (Christoffer Dall)
To: linux-arm-kernel@lists.infradead.org
Subject: [PULL 16/79] kvm: arm/arm64: Rework gpa callback handlers
Date: Sun, 23 Apr 2017 19:08:26 +0200	[thread overview]
Message-ID: <20170423170929.27334-17-cdall@linaro.org> (raw)
In-Reply-To: <20170423170929.27334-1-cdall@linaro.org>

From: Suzuki K Poulose <suzuki.poulose@arm.com>

In order to perform an operation on a gpa range, we currently iterate
over each page in a user memory slot for the given range. This is
inefficient while dealing with a big range (e.g, a VMA), especially
while unmaping a range. At present, with stage2 unmap on a range with
a hugepage backed region, we clear the PMD when we unmap the first
page in the loop. The remaining iterations simply traverse the page table
down to the PMD level only to see that nothing is in there.

This patch reworks the code to invoke the callback handlers on the
biggest range possible within the memory slot to to reduce the number of
times the handler is called.

Cc: Marc Zyngier <marc.zyngier@arm.com>
Reviewed-by: Christoffer Dall <cdall@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Christoffer Dall <cdall@linaro.org>
---
 arch/arm/kvm/mmu.c | 31 +++++++++++++------------------
 1 file changed, 13 insertions(+), 18 deletions(-)

diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index 962616f..69554bd 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -1512,7 +1512,8 @@ static int handle_hva_to_gpa(struct kvm *kvm,
 			     unsigned long start,
 			     unsigned long end,
 			     int (*handler)(struct kvm *kvm,
-					    gpa_t gpa, void *data),
+					    gpa_t gpa, u64 size,
+					    void *data),
 			     void *data)
 {
 	struct kvm_memslots *slots;
@@ -1524,7 +1525,7 @@ static int handle_hva_to_gpa(struct kvm *kvm,
 	/* we only care about the pages that the guest sees */
 	kvm_for_each_memslot(memslot, slots) {
 		unsigned long hva_start, hva_end;
-		gfn_t gfn, gfn_end;
+		gfn_t gpa;
 
 		hva_start = max(start, memslot->userspace_addr);
 		hva_end = min(end, memslot->userspace_addr +
@@ -1532,25 +1533,16 @@ static int handle_hva_to_gpa(struct kvm *kvm,
 		if (hva_start >= hva_end)
 			continue;
 
-		/*
-		 * {gfn(page) | page intersects with [hva_start, hva_end)} =
-		 * {gfn_start, gfn_start+1, ..., gfn_end-1}.
-		 */
-		gfn = hva_to_gfn_memslot(hva_start, memslot);
-		gfn_end = hva_to_gfn_memslot(hva_end + PAGE_SIZE - 1, memslot);
-
-		for (; gfn < gfn_end; ++gfn) {
-			gpa_t gpa = gfn << PAGE_SHIFT;
-			ret |= handler(kvm, gpa, data);
-		}
+		gpa = hva_to_gfn_memslot(hva_start, memslot) << PAGE_SHIFT;
+		ret |= handler(kvm, gpa, (u64)(hva_end - hva_start), data);
 	}
 
 	return ret;
 }
 
-static int kvm_unmap_hva_handler(struct kvm *kvm, gpa_t gpa, void *data)
+static int kvm_unmap_hva_handler(struct kvm *kvm, gpa_t gpa, u64 size, void *data)
 {
-	unmap_stage2_range(kvm, gpa, PAGE_SIZE);
+	unmap_stage2_range(kvm, gpa, size);
 	return 0;
 }
 
@@ -1577,10 +1569,11 @@ int kvm_unmap_hva_range(struct kvm *kvm,
 	return 0;
 }
 
-static int kvm_set_spte_handler(struct kvm *kvm, gpa_t gpa, void *data)
+static int kvm_set_spte_handler(struct kvm *kvm, gpa_t gpa, u64 size, void *data)
 {
 	pte_t *pte = (pte_t *)data;
 
+	WARN_ON(size != PAGE_SIZE);
 	/*
 	 * We can always call stage2_set_pte with KVM_S2PTE_FLAG_LOGGING_ACTIVE
 	 * flag clear because MMU notifiers will have unmapped a huge PMD before
@@ -1606,11 +1599,12 @@ void kvm_set_spte_hva(struct kvm *kvm, unsigned long hva, pte_t pte)
 	handle_hva_to_gpa(kvm, hva, end, &kvm_set_spte_handler, &stage2_pte);
 }
 
-static int kvm_age_hva_handler(struct kvm *kvm, gpa_t gpa, void *data)
+static int kvm_age_hva_handler(struct kvm *kvm, gpa_t gpa, u64 size, void *data)
 {
 	pmd_t *pmd;
 	pte_t *pte;
 
+	WARN_ON(size != PAGE_SIZE && size != PMD_SIZE);
 	pmd = stage2_get_pmd(kvm, NULL, gpa);
 	if (!pmd || pmd_none(*pmd))	/* Nothing there */
 		return 0;
@@ -1625,11 +1619,12 @@ static int kvm_age_hva_handler(struct kvm *kvm, gpa_t gpa, void *data)
 	return stage2_ptep_test_and_clear_young(pte);
 }
 
-static int kvm_test_age_hva_handler(struct kvm *kvm, gpa_t gpa, void *data)
+static int kvm_test_age_hva_handler(struct kvm *kvm, gpa_t gpa, u64 size, void *data)
 {
 	pmd_t *pmd;
 	pte_t *pte;
 
+	WARN_ON(size != PAGE_SIZE && size != PMD_SIZE);
 	pmd = stage2_get_pmd(kvm, NULL, gpa);
 	if (!pmd || pmd_none(*pmd))	/* Nothing there */
 		return 0;
-- 
2.9.0

  parent reply	other threads:[~2017-04-23 17:08 UTC|newest]

Thread overview: 81+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-04-23 17:08 [PULL 00/79] KVM/ARM Changes for v4.12 Christoffer Dall
2017-04-23 17:08 ` [PULL 01/79] arm64: sysreg: sort by encoding Christoffer Dall
2017-04-23 17:08 ` [PULL 02/79] arm64: sysreg: add debug system registers Christoffer Dall
2017-04-23 17:08 ` [PULL 03/79] arm64: sysreg: add performance monitor registers Christoffer Dall
2017-04-23 17:08 ` [PULL 04/79] arm64: sysreg: subsume GICv3 sysreg definitions Christoffer Dall
2017-04-23 17:08 ` [PULL 05/79] arm64: sysreg: add physical timer registers Christoffer Dall
2017-04-23 17:08 ` [PULL 06/79] arm64: sysreg: add register encodings used by KVM Christoffer Dall
2017-04-23 17:08 ` [PULL 07/79] arm64: sysreg: add Set/Way sys encodings Christoffer Dall
2017-04-23 17:08 ` [PULL 08/79] KVM: arm64: add SYS_DESC() Christoffer Dall
2017-04-23 17:08 ` [PULL 09/79] KVM: arm64: Use common debug sysreg definitions Christoffer Dall
2017-04-23 17:08 ` [PULL 10/79] KVM: arm64: Use common performance monitor " Christoffer Dall
2017-04-23 17:08 ` [PULL 11/79] KVM: arm64: Use common GICv3 " Christoffer Dall
2017-04-23 17:08 ` [PULL 12/79] KVM: arm64: Use common physical timer " Christoffer Dall
2017-04-23 17:08 ` [PULL 13/79] KVM: arm64: use common invariant " Christoffer Dall
2017-04-23 17:08 ` [PULL 14/79] KVM: arm64: Use common " Christoffer Dall
2017-04-23 17:08 ` [PULL 15/79] KVM: arm64: Use common Set/Way sys definitions Christoffer Dall
2017-04-23 17:08 ` Christoffer Dall [this message]
2017-04-23 17:08 ` [PULL 17/79] KVM: arm/arm64: vgic: Defer touching GICH_VMCR to vcpu_load/put Christoffer Dall
2017-04-23 17:08 ` [PULL 18/79] KVM: arm/arm64: vgic: Avoid flushing vgic state when there's no pending IRQ Christoffer Dall
2017-04-23 17:08 ` [PULL 19/79] KVM: arm/arm64: vgic: Get rid of live_lrs Christoffer Dall
2017-04-23 17:08 ` [PULL 20/79] KVM: arm/arm64: vgic: Only set underflow when actually out of LRs Christoffer Dall
2017-04-23 17:08 ` [PULL 21/79] KVM: arm/arm64: vgic: Get rid of unnecessary process_maintenance operation Christoffer Dall
2017-04-23 17:08 ` [PULL 22/79] KVM: arm/arm64: vgic: Get rid of unnecessary save_maint_int_state Christoffer Dall
2017-04-23 17:08 ` [PULL 23/79] KVM: arm/arm64: vgic: Get rid of MISR and EISR fields Christoffer Dall
2017-04-23 17:08 ` [PULL 24/79] KVM: arm/arm64: vgic: Implement early VGIC init functionality Christoffer Dall
2017-04-23 17:08 ` [PULL 25/79] KVM: arm/arm64: vgic: Don't check vgic_initialized in sync/flush Christoffer Dall
2017-04-23 17:08 ` [PULL 26/79] KVM: arm/arm64: vgic: Improve sync_hwstate performance Christoffer Dall
2017-04-23 17:08 ` [PULL 27/79] arm64: KVM: PMU: Refactor pmu_*_el0_disabled Christoffer Dall
2017-04-23 17:08 ` [PULL 28/79] arm64: KVM: PMU: Inject UNDEF exception on illegal register access Christoffer Dall
2017-04-23 17:08 ` [PULL 29/79] arm64: KVM: PMU: Inject UNDEF on non-privileged accesses Christoffer Dall
2017-04-23 17:08 ` [PULL 30/79] arm64: KVM: Make unexpected reads from WO registers inject an undef Christoffer Dall
2017-04-23 17:08 ` [PULL 31/79] arm64: KVM: PMU: Inject UNDEF on read access to PMSWINC_EL0 Christoffer Dall
2017-04-23 17:08 ` [PULL 32/79] arm64: KVM: Treat sysreg accessors returning false as successful Christoffer Dall
2017-04-23 17:08 ` [PULL 33/79] arm64: KVM: Do not corrupt registers on failed 64bit CP read Christoffer Dall
2017-04-23 17:08 ` [PULL 34/79] arm: KVM: Make unexpected register accesses inject an undef Christoffer Dall
2017-04-23 17:08 ` [PULL 35/79] arm: KVM: Treat CP15 accessors returning false as successful Christoffer Dall
2017-04-23 17:08 ` [PULL 36/79] arm64: hyp-stub: Stop pointlessly clobbering lr Christoffer Dall
2017-04-23 17:08 ` [PULL 37/79] arm64: KVM: Move lr save/restore to do_el2_call Christoffer Dall
2017-04-23 17:08 ` [PULL 38/79] arm64: hyp-stub: Don't save lr in the EL1 code Christoffer Dall
2017-04-23 17:08 ` [PULL 39/79] arm64: hyp-stub: Define a return value for failed stub calls Christoffer Dall
2017-04-23 17:08 ` [PULL 40/79] arm64: hyp-stub: Update documentation in asm/virt.h Christoffer Dall
2017-04-23 17:08 ` [PULL 41/79] arm64: hyp-stub: Implement HVC_RESET_VECTORS stub hypercall Christoffer Dall
2017-04-23 17:08 ` [PULL 42/79] arm64: KVM: Implement HVC_RESET_VECTORS stub hypercall in the init code Christoffer Dall
2017-04-23 17:08 ` [PULL 43/79] arm64: KVM: Implement HVC_GET_VECTORS " Christoffer Dall
2017-04-23 17:08 ` [PULL 44/79] arm64: KVM: Allow the main HYP code to use the init hyp stub implementation Christoffer Dall
2017-04-23 17:08 ` [PULL 45/79] arm64: KVM: Convert __cpu_reset_hyp_mode to using __hyp_reset_vectors Christoffer Dall
2017-04-23 17:08 ` [PULL 46/79] arm64: KVM: Implement HVC_SOFT_RESTART in the init code Christoffer Dall
2017-04-23 17:08 ` [PULL 47/79] ARM: hyp-stub: improve ABI Christoffer Dall
2017-04-23 17:08 ` [PULL 48/79] ARM: soft-reboot into same mode that we entered the kernel Christoffer Dall
2017-04-23 17:08 ` [PULL 49/79] ARM: KVM: Convert KVM to use HVC_GET_VECTORS Christoffer Dall
2017-04-23 17:09 ` [PULL 50/79] ARM: Update cpu_v7_reset documentation Christoffer Dall
2017-04-23 17:09 ` [PULL 51/79] ARM: hyp-stub: Use r1 for the soft-restart address Christoffer Dall
2017-04-23 17:09 ` [PULL 52/79] ARM: Expose the VA/IDMAP offset Christoffer Dall
2017-04-23 17:09 ` [PULL 53/79] ARM: hyp-stub: Define a return value for failed stub calls Christoffer Dall
2017-04-23 17:09 ` [PULL 54/79] ARM: hyp-stub: Implement HVC_RESET_VECTORS stub hypercall Christoffer Dall
2017-04-23 17:09 ` [PULL 55/79] ARM: KVM: Implement HVC_RESET_VECTORS stub hypercall in the init code Christoffer Dall
2017-04-23 17:09 ` [PULL 56/79] ARM: KVM: Implement HVC_GET_VECTORS " Christoffer Dall
2017-04-23 17:09 ` [PULL 57/79] ARM: KVM: Allow the main HYP code to use the init hyp stub implementation Christoffer Dall
2017-04-23 17:09 ` [PULL 58/79] ARM: KVM: Convert __cpu_reset_hyp_mode to using __hyp_reset_vectors Christoffer Dall
2017-04-23 17:09 ` [PULL 59/79] ARM: KVM: Implement HVC_SOFT_RESTART in the init code Christoffer Dall
2017-04-23 17:09 ` [PULL 60/79] ARM: KVM: Gracefully handle hyp-stubs being restored from under our feet Christoffer Dall
2017-04-23 17:09 ` [PULL 61/79] arm/arm64: KVM: Use __hyp_reset_vectors() directly Christoffer Dall
2017-04-23 17:09 ` [PULL 62/79] arm/arm64: KVM: Remove kvm_get_idmap_start Christoffer Dall
2017-04-23 17:09 ` [PULL 63/79] arm/arm64: KVM: Use HVC_RESET_VECTORS to reinit HYP mode Christoffer Dall
2017-04-23 17:09 ` [PULL 64/79] ARM: decompressor: Remove __hyp_get_vectors usage Christoffer Dall
2017-04-23 17:09 ` [PULL 65/79] ARM: hyp-stub/KVM: Kill __hyp_get_vectors Christoffer Dall
2017-04-23 17:09 ` [PULL 66/79] arm64: " Christoffer Dall
2017-04-23 17:09 ` [PULL 67/79] arm64: hyp-stub: Zero x0 on successful stub handling Christoffer Dall
2017-04-23 17:09 ` [PULL 68/79] ARM: hyp-stub: Zero r0 " Christoffer Dall
2017-04-23 17:09 ` [PULL 69/79] arm/arm64: Add hyp-stub API documentation Christoffer Dall
2017-04-23 17:09 ` [PULL 70/79] KVM: arm/arm64: Cleanup the arch timer code's irqchip checking Christoffer Dall
2017-04-23 17:09 ` [PULL 71/79] KVM: arm/arm64: Add ARM user space interrupt signaling ABI Christoffer Dall
2017-04-23 17:09 ` [PULL 72/79] KVM: arm/arm64: Support arch timers with a userspace gic Christoffer Dall
2017-04-23 17:09 ` [PULL 73/79] KVM: arm/arm64: Report PMU overflow interrupts to userspace irqchip Christoffer Dall
2017-04-23 17:09 ` [PULL 74/79] KVM: arm/arm64: Advertise support for KVM_CAP_ARM_USER_IRQ Christoffer Dall
2017-04-23 17:09 ` [PULL 75/79] KVM: arm/arm64: fix races in kvm_psci_vcpu_on Christoffer Dall
2017-04-23 17:09 ` [PULL 76/79] KVM: arm/arm64: vgic-v3: De-optimize VMCR save/restore when emulating a GICv2 Christoffer Dall
2017-04-23 17:09 ` [PULL 77/79] KVM: arm/arm64: vgic-v3: Fix off-by-one LR access Christoffer Dall
2017-04-23 17:09 ` [PULL 78/79] ARM: hyp-stub: Fix Thumb-2 compilation Christoffer Dall
2017-04-23 17:09 ` [PULL 79/79] ARM: KVM: Fix idmap stub entry when running Thumb-2 code Christoffer Dall
2017-04-27 15:34 ` [PULL 00/79] KVM/ARM Changes for v4.12 Paolo Bonzini

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170423170929.27334-17-cdall@linaro.org \
    --to=cdall@linaro.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).