From: "Philippe Mathieu-Daudé" <philmd@linaro.org>
To: qemu-devel@nongnu.org
Cc: "Paolo Bonzini" <pbonzini@redhat.com>,
	"Richard Henderson" <richard.henderson@linaro.org>,
	"Mohamed Mediouni" <mohamed@unpredictable.fr>,
	"Stefan Hajnoczi" <stefanha@redhat.com>,
	"Alexander Graf" <agraf@csgraf.de>,
	"Peter Collingbourne" <pcc@google.com>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>,
	"Cameron Esfahani" <dirty@apple.com>,
	qemu-arm@nongnu.org, "Mads Ynddal" <mads@ynddal.dk>,
	"Phil Dennis-Jordan" <phil@philjordan.eu>,
	"Peter Maydell" <peter.maydell@linaro.org>,
	"Roman Bolshakov" <rbolshakov@ddn.com>
Subject: [PATCH v2 42/58] target/arm/hvf: Implement dirty page tracking
Date: Thu, 23 Oct 2025 15:06:18 +0200	[thread overview]
Message-ID: <20251023130625.9157-1-philmd@linaro.org> (raw)
In-Reply-To: <20251023114638.5667-1-philmd@linaro.org>
From: Richard Henderson <richard.henderson@linaro.org>
Notice writes to pages which are being monitored.  Mark the page dirty,
re-enable writes, and retry the instruction without emulation.
Assert the fault is not from a stage1 page table walk.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
---
 target/arm/hvf/hvf.c | 52 ++++++++++++++++++++++++++++++++++++--------
 1 file changed, 43 insertions(+), 9 deletions(-)
diff --git a/target/arm/hvf/hvf.c b/target/arm/hvf/hvf.c
index 77998ab90a6..cf4746c344a 100644
--- a/target/arm/hvf/hvf.c
+++ b/target/arm/hvf/hvf.c
@@ -37,6 +37,7 @@
 #include "target/arm/trace.h"
 #include "trace.h"
 #include "migration/vmstate.h"
+#include "exec/target_page.h"
 
 #include "gdbstub/enums.h"
 
@@ -1880,9 +1881,10 @@ static int hvf_handle_exception(CPUState *cpu, hv_vcpu_exit_exception_t *excp)
         uint32_t srt = (syndrome >> 16) & 0x1f;
         uint32_t cm = (syndrome >> 8) & 0x1;
         uint64_t val = 0;
+        uint64_t ipa = excp->physical_address;
+        AddressSpace *as = cpu_get_address_space(cpu, ARMASIdx_NS);
 
-        trace_hvf_data_abort(excp->virtual_address,
-                             excp->physical_address, isv,
+        trace_hvf_data_abort(excp->virtual_address, ipa, isv,
                              iswrite, s1ptw, len, srt);
 
         if (cm) {
@@ -1891,23 +1893,55 @@ static int hvf_handle_exception(CPUState *cpu, hv_vcpu_exit_exception_t *excp)
             break;
         }
 
+        /* Handle dirty page logging for ram. */
+        if (iswrite) {
+            hwaddr xlat;
+            MemoryRegion *mr = address_space_translate(as, ipa, &xlat,
+                                                       NULL, true,
+                                                       MEMTXATTRS_UNSPECIFIED);
+            if (memory_region_is_ram(mr)) {
+                uint64_t ipa_page = ipa & TARGET_PAGE_MASK;
+
+                /* TODO: Inject exception to the guest. */
+                assert(!mr->readonly);
+
+                if (memory_region_get_dirty_log_mask(mr)) {
+                    memory_region_set_dirty(mr, ipa_page + xlat,
+                                            TARGET_PAGE_SIZE);
+                    hvf_unprotect_dirty_range(ipa_page, TARGET_PAGE_SIZE);
+                }
+
+                /* Retry with page writes enabled. */
+                break;
+            }
+        }
+
+        /*
+         * TODO: If s1ptw, this is an error in the guest os page tables.
+         * Inject the exception into the guest.
+         */
+        assert(!s1ptw);
+
+        /*
+         * TODO: ISV will be 0 for SIMD or SVE accesses.
+         * Inject the exception into the guest.
+         */
         assert(isv);
 
+        /*
+         * Emulate MMIO.
+         * TODO: Inject faults for errors.
+         */
         if (iswrite) {
             val = hvf_get_reg(cpu, srt);
-            address_space_write(&address_space_memory,
-                                excp->physical_address,
-                                MEMTXATTRS_UNSPECIFIED, &val, len);
+            address_space_write(as, ipa, MEMTXATTRS_UNSPECIFIED, &val, len);
         } else {
-            address_space_read(&address_space_memory,
-                               excp->physical_address,
-                               MEMTXATTRS_UNSPECIFIED, &val, len);
+            address_space_read(as, ipa, MEMTXATTRS_UNSPECIFIED, &val, len);
             if (sse) {
                 val = sextract64(val, 0, len * 8);
             }
             hvf_set_reg(cpu, srt, val);
         }
-
         advance_pc = true;
         break;
     }
-- 
2.51.0
next prev parent reply	other threads:[~2025-10-23 13:07 UTC|newest]
Thread overview: 78+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-23 11:45 [PATCH v2 00/58] target/arm/hvf: Consolidate Philippe Mathieu-Daudé
2025-10-23 11:45 ` [PATCH v2 01/58] target/arm/hvf: Release memory allocated by hv_vcpu_config_create() Philippe Mathieu-Daudé
2025-10-23 11:45 ` [PATCH v2 02/58] target/arm/hvf: Trace vCPU KICK events Philippe Mathieu-Daudé
2025-10-25 17:01   ` Richard Henderson
2025-10-23 11:45 ` [PATCH v2 03/58] target/arm/hvf: Check hv_vcpus_exit() returned value Philippe Mathieu-Daudé
2025-10-25 17:02   ` Richard Henderson
2025-10-23 11:45 ` [PATCH v2 04/58] target/arm/hvf: Check hv_vcpu_set_vtimer_mask() " Philippe Mathieu-Daudé
2025-10-23 11:45 ` [PATCH v2 05/58] accel/hvf: Rename hvf_vcpu_exec() -> hvf_arch_vcpu_exec() Philippe Mathieu-Daudé
2025-10-23 11:45 ` [PATCH v2 06/58] accel/hvf: Rename hvf_put|get_registers -> hvf_arch_put|get_registers Philippe Mathieu-Daudé
2025-10-23 11:45 ` [PATCH v2 07/58] target/arm/hvf: Mention flush_cpu_state() must run on vCPU thread Philippe Mathieu-Daudé
2025-10-23 11:45 ` [PATCH v2 08/58] accel/hvf: Mention hvf_arch_init_vcpu() " Philippe Mathieu-Daudé
2025-10-23 11:45 ` [PATCH v2 09/58] target/arm/hvf: Mention hvf_sync_vtimer() " Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 30/58] target/arm: Re-use arm_is_psci_call() in HVF Philippe Mathieu-Daudé
2025-10-25 17:28   ` Richard Henderson
2025-10-23 11:52 ` [PATCH v2 31/58] target/arm: Share ARM_PSCI_CALL trace event between TCG and HVF Philippe Mathieu-Daudé
2025-10-25 17:29   ` Richard Henderson
2025-10-23 11:52 ` [PATCH v2 32/58] target/arm/hvf/hvf: Document $pc adjustment in HVF & SMC Philippe Mathieu-Daudé
2025-10-25 17:30   ` Richard Henderson
2025-10-23 11:52 ` [PATCH v2 33/58] accel/hvf: Trace prefetch abort Philippe Mathieu-Daudé
2025-10-25 17:31   ` Richard Henderson
2025-10-23 11:52 ` [PATCH v2 34/58] accel/hvf: Create hvf_protect_clean_range, hvf_unprotect_dirty_range Philippe Mathieu-Daudé
2025-10-23 14:18   ` Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 35/58] target/i386/hvf: Use hvf_unprotect_page Philippe Mathieu-Daudé
2025-10-23 14:19   ` Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 36/58] target/i386/hvf: Use address_space_translate in ept_emulation_fault Philippe Mathieu-Daudé
2025-10-23 14:39   ` Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 37/58] accel/hvf: Simplify hvf_log_* Philippe Mathieu-Daudé
2025-10-23 14:40   ` Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 38/58] accel/hvf: Move hvf_log_sync to hvf_log_clear Philippe Mathieu-Daudé
2025-10-23 14:19   ` Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 39/58] accel/hvf: Simplify hvf_set_phys_mem Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 40/58] accel/hvf: Drop hvf_slot and hvf_find_overlap_slot Philippe Mathieu-Daudé
2025-10-23 14:20   ` Philippe Mathieu-Daudé
2025-10-23 11:52 ` [PATCH v2 41/58] accel/hvf: Remove mac_slots Philippe Mathieu-Daudé
2025-10-23 14:20   ` Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 50/58] accel/hvf: Sync CNTV_CTL_EL0 & CNTV_CVAL_EL0 Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 51/58] accel/hvf: Model PhysTimer register Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 52/58] accel: Introduce AccelOpsClass::cpu_target_realize() hook Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 53/58] accel/hvf: Add hvf_arch_cpu_realize() stubs Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 54/58] target/arm: Create GTimers *after* features finalized / accel realized Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 55/58] target/arm/hvf: Really set Generic Timer counter frequency Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 56/58] accel/hvf: Do not abort in hvf_arm_get_*_ipa_bit_size() Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 57/58] hw/arm/virt: Warn when HVF doesn't report IPA bit length Philippe Mathieu-Daudé
2025-10-23 12:31 ` [PATCH v2 58/58] target/arm: Only allow disabling NEON when using TCG Philippe Mathieu-Daudé
2025-10-25 17:58   ` Richard Henderson
2025-10-23 13:06 ` Philippe Mathieu-Daudé [this message]
2025-10-23 14:43   ` [PATCH v2 42/58] target/arm/hvf: Implement dirty page tracking Philippe Mathieu-Daudé
2025-10-25 18:25   ` Richard Henderson
2025-10-23 13:06 ` [PATCH v2 43/58] accel/hvf: Enforce host alignment when calling hv_vm_protect() Philippe Mathieu-Daudé
2025-10-23 13:06 ` [PATCH v2 44/58] accel/hvf: Have WFI returns if !cpu_has_work Philippe Mathieu-Daudé
2025-10-25 18:29   ` Richard Henderson
2025-10-23 13:06 ` [PATCH v2 45/58] accel/hvf: Implement WFI without using pselect() Philippe Mathieu-Daudé
2025-10-23 13:06 ` [PATCH v2 46/58] accel/hvf: Have PSCI CPU_SUSPEND halt the vCPU Philippe Mathieu-Daudé
2025-10-25 19:27   ` Richard Henderson
2025-10-23 13:06 ` [PATCH v2 47/58] accel/hvf: Introduce hvf_arch_cpu_synchronize_[pre/post]exec() hooks Philippe Mathieu-Daudé
2025-10-23 13:06 ` [PATCH v2 48/58] target/i386/hvf: Flush vCPU registers once before vcpu_exec() loop Philippe Mathieu-Daudé
2025-10-23 13:06 ` [PATCH v2 49/58] target/arm/hvf: " Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 10/58] target/arm/hvf: Mention hvf_arch_set_traps() must run on vCPU thread Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 11/58] accel/hvf: Mention hvf_arch_update_guest_debug() must run on vCPU Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 12/58] target/arm/hvf: Mention hvf_inject_interrupts() must run on vCPU thread Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 13/58] accel/hvf: Implement hvf_arch_vcpu_destroy() Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 14/58] target/arm/hvf: Hardcode Apple MIDR Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 15/58] target/arm/hvf: Simplify hvf_arm_get_host_cpu_features() Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 16/58] target/arm/hvf: switch hvf_arm_get_host_cpu_features to not create a vCPU Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 17/58] target/arm/hvf: Factor hvf_handle_exception() out Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 18/58] target/i386/hvf: Factor hvf_handle_vmexit() out Philippe Mathieu-Daudé
2025-10-23 14:13 ` [PATCH v2 19/58] target/arm/hvf: " Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 20/58] target/arm/hvf: Keep calling hv_vcpu_run() in loop Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 21/58] cpus: Trace cpu_exec_start() and cpu_exec_end() calls Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 22/58] accel/hvf: Guard hv_vcpu_run() between cpu_exec_start/end() calls Philippe Mathieu-Daudé
2025-10-23 14:36   ` Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 23/58] target/arm: Call aarch64_add_pauth_properties() once in host_initfn() Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 24/58] accel/hvf: Restrict ARM specific fields of AccelCPUState Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 25/58] target/arm: Rename init_cpreg_list() -> arm_init_cpreg_list() Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 26/58] target/arm: Better describe PMU depends on TCG or HVF Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 27/58] target/arm/hvf: Rename 'vgic' -> 'emu_reginfo' in trace events Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 28/58] target/arm/hvf: Emulate PMU registers Philippe Mathieu-Daudé
2025-10-23 14:30 ` [PATCH v2 29/58] target/arm/hvf: Emulate Monitor Debug registers Philippe Mathieu-Daudé
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=20251023130625.9157-1-philmd@linaro.org \
    --to=philmd@linaro.org \
    --cc=agraf@csgraf.de \
    --cc=dirty@apple.com \
    --cc=mads@ynddal.dk \
    --cc=mohamed@unpredictable.fr \
    --cc=pbonzini@redhat.com \
    --cc=pcc@google.com \
    --cc=peter.maydell@linaro.org \
    --cc=phil@philjordan.eu \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=rbolshakov@ddn.com \
    --cc=richard.henderson@linaro.org \
    --cc=stefanha@redhat.com \
    /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).