On Thu, 13 Nov 2025 03:25:27 +1000 Gavin Shan <[email protected]> wrote:
> In the combination of 64KiB host and 4KiB guest, a problematic host > page affects 16x guest pages. Those 16x guest pages are most likely > owned by separate threads and accessed by the threads in parallel. > It means 16x memory errors can be raised at once. However, we're > unable to handle this situation because the only error source has > one read acknowledgement register in current design. QEMU has to > crash in the following path due to the previously delivered error > isn't acknowledged by the guest on attempt to deliver another error. > > kvm_vcpu_thread_fn > kvm_cpu_exec > kvm_arch_on_sigbus_vcpu > kvm_cpu_synchronize_state > acpi_ghes_memory_errors > abort > > This series fixes the issue by sending 16x consective CPER errors > which are contained in a single GHES error block. > > PATCH[1-4] Increases GHES raw data maximal length from 1KiB to 4KiB > PATCH[5] Supports multiple error records in a single error block > PATCH[6-7] Improves the error handling in the error delivery path > PATCH[8] Sends 16x consective CPERs in a single block if needed > Hi Gavin, Just a quick head's up to say we've had some internal discussions around the kernel handling of broader address masks in CPER and think it is probably broken. Rectifying that may at least simplify what is needed on the QEMU side of things and maybe even handle much larger blocks (2M and larger). Will keep everyone informed of how we get on with resolving that. Jonathan > Changelog > ========= > v4: > * v3: https://lists.nongnu.org/archive/html/qemu-arm/2025-11/msg00264.html > * Pick r-b tags from Jonathan > * Add compat property 'x-error-block-size' for migration (Igor) > * Code and commit log improvements (Igor/Jonathan) > * Use error_fatal in the memory error delivery path (Markus) > * Use APIs from registerfields.h (Philippe) > v3: > * v2: https://lists.nongnu.org/archive/html/qemu-arm/2025-10/msg00372.html > * Code and changelog improvements (Jonathan) > * Fixed GHES error block status field and improved error > handling in the error delivery path (Igor) > * Fixed ACPI HEST table and document (Mauro) > v2: > * v1: https://lists.nongnu.org/archive/html/qemu-arm/2025-02/msg00897.html > * Send 16x memory errors for the specific case (Jonathan) > > Gavin Shan (8): > acpi/ghes: Make GHES max raw data length dynamic > tests/qtest/bios-tables-test: Prepare for changes in the HEST table > acpi/ghes: Increase GHES raw data maximal length to 4KiB > tests/qtest/bios-tables-test: Update HEST table > acpi/ghes: Extend acpi_ghes_memory_errors() for multiple CPERs > acpi/ghes: Bail early on error from get_ghes_source_offsets() > acpi/ghes: Use error_fatal in acpi_ghes_memory_errors() > target/arm/kvm: Support multiple memory CPERs injection > > docs/specs/acpi_hest_ghes.rst | 6 +- > hw/acpi/generic_event_device.c | 2 + > hw/acpi/ghes-stub.c | 7 +- > hw/acpi/ghes.c | 127 +++++++++++++++++------------- > hw/core/machine.c | 1 + > include/hw/acpi/ghes.h | 8 +- > target/arm/kvm.c | 72 +++++++++++++++-- > tests/data/acpi/aarch64/virt/HEST | Bin 224 -> 224 bytes > 8 files changed, 153 insertions(+), 70 deletions(-) >
