From: "Philippe Mathieu-Daudé" <philmd@redhat.com>
To: Samuel Ortiz <sameo@linux.intel.com>, qemu-devel@nongnu.org
Cc: Yang Zhong <yang.zhong@intel.com>,
Peter Maydell <peter.maydell@linaro.org>,
Eduardo Habkost <ehabkost@redhat.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
Paolo Bonzini <pbonzini@redhat.com>,
Shannon Zhao <zhaoshenglong@huawei.com>,
Igor Mammedov <imammedo@redhat.com>,
"open list:ARM ACPI Subsystem" <qemu-arm@nongnu.org>,
Richard Henderson <rth@twiddle.net>
Subject: Re: [Qemu-devel] [PATCH v3 13/19] hw: acpi: Export the SRAT AML build API
Date: Mon, 29 Oct 2018 20:40:00 +0100 [thread overview]
Message-ID: <e7cc3caf-3ee6-c9a8-0c2d-482f06a02019@redhat.com> (raw)
In-Reply-To: <20181029170159.3801-14-sameo@linux.intel.com>
On 29/10/18 18:01, Samuel Ortiz wrote:
> From: Yang Zhong <yang.zhong@intel.com>
>
> The SRAT ACPI table is not x86 specific and will be needed for the
> Hardware-reduced ACPI implementation. So we should export it through the
> architecture independent hw/acpi folder.
>
> Also, now that the generic build_srat() API is exported, we have to
> rename the ARM static one in order to avoid build time conflicts.
>
> Signed-off-by: Yang Zhong <yang.zhong@intel.com>
> ---
> hw/acpi/aml-build.c | 130 ++++++++++++++++++++++++++++++++++++
> hw/arm/virt-acpi-build.c | 4 +-
> hw/i386/acpi-build.c | 129 -----------------------------------
> include/hw/acpi/aml-build.h | 3 +
> 4 files changed, 135 insertions(+), 131 deletions(-)
>
> diff --git a/hw/acpi/aml-build.c b/hw/acpi/aml-build.c
> index 2110e18799..c3f652a68f 100644
> --- a/hw/acpi/aml-build.c
> +++ b/hw/acpi/aml-build.c
> @@ -22,6 +22,7 @@
> #include "qemu/osdep.h"
> #include <glib/gprintf.h>
> #include "hw/acpi/aml-build.h"
> +#include "hw/mem/memory-device.h"
> #include "qemu/bswap.h"
> #include "qemu/bitops.h"
> #include "sysemu/numa.h"
> @@ -2459,6 +2460,135 @@ void build_srat_memory(AcpiSratMemoryAffinity *numamem, uint64_t base,
> numamem->range_length = cpu_to_le64(len);
> }
>
> +#define HOLE_640K_START (640 * KiB)
> +#define HOLE_640K_END (1 * MiB)
> +
> +void
> +build_srat(GArray *table_data, BIOSLinker *linker,
> + MachineState *machine, AcpiConfiguration *conf)
> +{
> + AcpiSystemResourceAffinityTable *srat;
> + AcpiSratMemoryAffinity *numamem;
> +
> + int i;
> + int srat_start, numa_start, slots;
> + uint64_t mem_len, mem_base, next_base;
> + MachineClass *mc = MACHINE_GET_CLASS(machine);
> + const CPUArchIdList *apic_ids = mc->possible_cpu_arch_ids(machine);
> + ram_addr_t hotplugabble_address_space_size =
> + object_property_get_int(OBJECT(machine), MEMORY_DEVICE_REGION_SIZE,
> + NULL);
> +
> + srat_start = table_data->len;
> +
> + srat = acpi_data_push(table_data, sizeof *srat);
> + srat->reserved1 = cpu_to_le32(1);
> +
> + for (i = 0; i < apic_ids->len; i++) {
> + int node_id = apic_ids->cpus[i].props.node_id;
> + uint32_t apic_id = apic_ids->cpus[i].arch_id;
> +
> + if (apic_id < 255) {
> + AcpiSratProcessorAffinity *core;
> +
> + core = acpi_data_push(table_data, sizeof *core);
> + core->type = ACPI_SRAT_PROCESSOR_APIC;
> + core->length = sizeof(*core);
> + core->local_apic_id = apic_id;
> + core->proximity_lo = node_id;
> + memset(core->proximity_hi, 0, 3);
> + core->local_sapic_eid = 0;
> + core->flags = cpu_to_le32(1);
> + } else {
> + AcpiSratProcessorX2ApicAffinity *core;
> +
> + core = acpi_data_push(table_data, sizeof *core);
> + core->type = ACPI_SRAT_PROCESSOR_x2APIC;
> + core->length = sizeof(*core);
> + core->x2apic_id = cpu_to_le32(apic_id);
> + core->proximity_domain = cpu_to_le32(node_id);
> + core->flags = cpu_to_le32(1);
> + }
> + }
> +
> +
> + /* the memory map is a bit tricky, it contains at least one hole
> + * from 640k-1M and possibly another one from 3.5G-4G.
> + */
> + next_base = 0;
> + numa_start = table_data->len;
> +
> + for (i = 1; i < conf->numa_nodes + 1; ++i) {
> + mem_base = next_base;
> + mem_len = conf->node_mem[i - 1];
> + next_base = mem_base + mem_len;
> +
> + /* Cut out the 640K hole */
It seems we are now using x86 specific code for a supposed arch agnostic
file. I'm not very comfortable with this change.
> + if (mem_base <= HOLE_640K_START &&
> + next_base > HOLE_640K_START) {
> + mem_len -= next_base - HOLE_640K_START;
> + if (mem_len > 0) {
> + numamem = acpi_data_push(table_data, sizeof *numamem);
> + build_srat_memory(numamem, mem_base, mem_len, i - 1,
> + MEM_AFFINITY_ENABLED);
> + }
> +
> + /* Check for the rare case: 640K < RAM < 1M */
> + if (next_base <= HOLE_640K_END) {
> + next_base = HOLE_640K_END;
> + continue;
> + }
> + mem_base = HOLE_640K_END;
> + mem_len = next_base - HOLE_640K_END;
> + }
> +
> + /* Cut out the ACPI_PCI hole */
> + if (mem_base <= conf->below_4g_mem_size &&
> + next_base > conf->below_4g_mem_size) {
> + mem_len -= next_base - conf->below_4g_mem_size;
> + if (mem_len > 0) {
> + numamem = acpi_data_push(table_data, sizeof *numamem);
> + build_srat_memory(numamem, mem_base, mem_len, i - 1,
> + MEM_AFFINITY_ENABLED);
> + }
> + mem_base = 1ULL << 32;
> + mem_len = next_base - conf->below_4g_mem_size;
> + next_base = mem_base + mem_len;
> + }
> +
> + if (mem_len > 0) {
> + numamem = acpi_data_push(table_data, sizeof *numamem);
> + build_srat_memory(numamem, mem_base, mem_len, i - 1,
> + MEM_AFFINITY_ENABLED);
> + }
> + }
> + slots = (table_data->len - numa_start) / sizeof *numamem;
> + for (; slots < conf->numa_nodes + 2; slots++) {
> + numamem = acpi_data_push(table_data, sizeof *numamem);
> + build_srat_memory(numamem, 0, 0, 0, MEM_AFFINITY_NOFLAGS);
> + }
> +
> + /*
> + * Entry is required for Windows to enable memory hotplug in OS
> + * and for Linux to enable SWIOTLB when booted with less than
> + * 4G of RAM. Windows works better if the entry sets proximity
> + * to the highest NUMA node in the machine.
> + * Memory devices may override proximity set by this entry,
> + * providing _PXM method if necessary.
> + */
> + if (hotplugabble_address_space_size) {
> + numamem = acpi_data_push(table_data, sizeof *numamem);
> + build_srat_memory(numamem, machine->device_memory->base,
> + hotplugabble_address_space_size, conf->numa_nodes - 1,
> + MEM_AFFINITY_HOTPLUGGABLE | MEM_AFFINITY_ENABLED);
> + }
> +
> + build_header(linker, table_data,
> + (void *)(table_data->data + srat_start),
> + "SRAT",
> + table_data->len - srat_start, 1, NULL, NULL);
> +}
> +
> /*
> * ACPI spec 5.2.17 System Locality Distance Information Table
> * (Revision 2.0 or later)
> diff --git a/hw/arm/virt-acpi-build.c b/hw/arm/virt-acpi-build.c
> index f9a60907f1..353aa55357 100644
> --- a/hw/arm/virt-acpi-build.c
> +++ b/hw/arm/virt-acpi-build.c
> @@ -469,7 +469,7 @@ build_spcr(GArray *table_data, BIOSLinker *linker, VirtMachineState *vms)
> }
>
> static void
> -build_srat(GArray *table_data, BIOSLinker *linker, VirtMachineState *vms)
> +virt_build_srat(GArray *table_data, BIOSLinker *linker, VirtMachineState *vms)
> {
> AcpiSystemResourceAffinityTable *srat;
> AcpiSratProcessorGiccAffinity *core;
> @@ -759,7 +759,7 @@ void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables)
>
> if (nb_numa_nodes > 0) {
> acpi_add_table(table_offsets, tables_blob);
> - build_srat(tables_blob, tables->linker, vms);
> + virt_build_srat(tables_blob, tables->linker, vms);
> if (have_numa_distance) {
> acpi_add_table(table_offsets, tables_blob);
> build_slit(tables_blob, tables->linker);
> diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
> index dfc02a8a85..5932dbe825 100644
> --- a/hw/i386/acpi-build.c
> +++ b/hw/i386/acpi-build.c
> @@ -1614,135 +1614,6 @@ build_tpm2(GArray *table_data, BIOSLinker *linker, GArray *tcpalog)
> (void *)tpm2_ptr, "TPM2", sizeof(*tpm2_ptr), 4, NULL, NULL);
> }
>
> -#define HOLE_640K_START (640 * KiB)
> -#define HOLE_640K_END (1 * MiB)
> -
> -static void
> -build_srat(GArray *table_data, BIOSLinker *linker,
> - MachineState *machine, AcpiConfiguration *conf)
> -{
> - AcpiSystemResourceAffinityTable *srat;
> - AcpiSratMemoryAffinity *numamem;
> -
> - int i;
> - int srat_start, numa_start, slots;
> - uint64_t mem_len, mem_base, next_base;
> - MachineClass *mc = MACHINE_GET_CLASS(machine);
> - const CPUArchIdList *apic_ids = mc->possible_cpu_arch_ids(machine);
> - ram_addr_t hotplugabble_address_space_size =
> - object_property_get_int(OBJECT(machine), MEMORY_DEVICE_REGION_SIZE,
> - NULL);
> -
> - srat_start = table_data->len;
> -
> - srat = acpi_data_push(table_data, sizeof *srat);
> - srat->reserved1 = cpu_to_le32(1);
> -
> - for (i = 0; i < apic_ids->len; i++) {
> - int node_id = apic_ids->cpus[i].props.node_id;
> - uint32_t apic_id = apic_ids->cpus[i].arch_id;
> -
> - if (apic_id < 255) {
> - AcpiSratProcessorAffinity *core;
> -
> - core = acpi_data_push(table_data, sizeof *core);
> - core->type = ACPI_SRAT_PROCESSOR_APIC;
> - core->length = sizeof(*core);
> - core->local_apic_id = apic_id;
> - core->proximity_lo = node_id;
> - memset(core->proximity_hi, 0, 3);
> - core->local_sapic_eid = 0;
> - core->flags = cpu_to_le32(1);
> - } else {
> - AcpiSratProcessorX2ApicAffinity *core;
> -
> - core = acpi_data_push(table_data, sizeof *core);
> - core->type = ACPI_SRAT_PROCESSOR_x2APIC;
> - core->length = sizeof(*core);
> - core->x2apic_id = cpu_to_le32(apic_id);
> - core->proximity_domain = cpu_to_le32(node_id);
> - core->flags = cpu_to_le32(1);
> - }
> - }
> -
> -
> - /* the memory map is a bit tricky, it contains at least one hole
> - * from 640k-1M and possibly another one from 3.5G-4G.
> - */
> - next_base = 0;
> - numa_start = table_data->len;
> -
> - for (i = 1; i < conf->numa_nodes + 1; ++i) {
> - mem_base = next_base;
> - mem_len = conf->node_mem[i - 1];
> - next_base = mem_base + mem_len;
> -
> - /* Cut out the 640K hole */
> - if (mem_base <= HOLE_640K_START &&
> - next_base > HOLE_640K_START) {
> - mem_len -= next_base - HOLE_640K_START;
> - if (mem_len > 0) {
> - numamem = acpi_data_push(table_data, sizeof *numamem);
> - build_srat_memory(numamem, mem_base, mem_len, i - 1,
> - MEM_AFFINITY_ENABLED);
> - }
> -
> - /* Check for the rare case: 640K < RAM < 1M */
> - if (next_base <= HOLE_640K_END) {
> - next_base = HOLE_640K_END;
> - continue;
> - }
> - mem_base = HOLE_640K_END;
> - mem_len = next_base - HOLE_640K_END;
> - }
> -
> - /* Cut out the ACPI_PCI hole */
> - if (mem_base <= conf->below_4g_mem_size &&
> - next_base > conf->below_4g_mem_size) {
> - mem_len -= next_base - conf->below_4g_mem_size;
> - if (mem_len > 0) {
> - numamem = acpi_data_push(table_data, sizeof *numamem);
> - build_srat_memory(numamem, mem_base, mem_len, i - 1,
> - MEM_AFFINITY_ENABLED);
> - }
> - mem_base = 1ULL << 32;
> - mem_len = next_base - conf->below_4g_mem_size;
> - next_base = mem_base + mem_len;
> - }
> -
> - if (mem_len > 0) {
> - numamem = acpi_data_push(table_data, sizeof *numamem);
> - build_srat_memory(numamem, mem_base, mem_len, i - 1,
> - MEM_AFFINITY_ENABLED);
> - }
> - }
> - slots = (table_data->len - numa_start) / sizeof *numamem;
> - for (; slots < conf->numa_nodes + 2; slots++) {
> - numamem = acpi_data_push(table_data, sizeof *numamem);
> - build_srat_memory(numamem, 0, 0, 0, MEM_AFFINITY_NOFLAGS);
> - }
> -
> - /*
> - * Entry is required for Windows to enable memory hotplug in OS
> - * and for Linux to enable SWIOTLB when booted with less than
> - * 4G of RAM. Windows works better if the entry sets proximity
> - * to the highest NUMA node in the machine.
> - * Memory devices may override proximity set by this entry,
> - * providing _PXM method if necessary.
> - */
> - if (hotplugabble_address_space_size) {
> - numamem = acpi_data_push(table_data, sizeof *numamem);
> - build_srat_memory(numamem, machine->device_memory->base,
> - hotplugabble_address_space_size, conf->numa_nodes - 1,
> - MEM_AFFINITY_HOTPLUGGABLE | MEM_AFFINITY_ENABLED);
> - }
> -
> - build_header(linker, table_data,
> - (void *)(table_data->data + srat_start),
> - "SRAT",
> - table_data->len - srat_start, 1, NULL, NULL);
> -}
> -
> /*
> * VT-d spec 8.1 DMA Remapping Reporting Structure
> * (version Oct. 2014 or later)
> diff --git a/include/hw/acpi/aml-build.h b/include/hw/acpi/aml-build.h
> index 1fabf58df2..654ce2ec26 100644
> --- a/include/hw/acpi/aml-build.h
> +++ b/include/hw/acpi/aml-build.h
> @@ -1,6 +1,7 @@
> #ifndef HW_ACPI_AML_BUILD_H
> #define HW_ACPI_AML_BUILD_H
>
> +#include "hw/acpi/acpi.h"
> #include "hw/acpi/acpi-defs.h"
> #include "hw/acpi/bios-linker-loader.h"
> #include "hw/pci/pcie_host.h"
> @@ -438,6 +439,8 @@ void
> build_xsdt(GArray *table_data, BIOSLinker *linker, GArray *table_offsets,
> const char *oem_id, const char *oem_table_id);
>
> +void build_srat(GArray *table_data, BIOSLinker *linker,
> + MachineState *machine, AcpiConfiguration *conf);
> int
> build_append_named_dword(GArray *array, const char *name_format, ...)
> GCC_FMT_ATTR(2, 3);
>
next prev parent reply other threads:[~2018-10-29 19:40 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-29 17:01 [Qemu-devel] [PATCH v3 00/19] ACPI reorganization for hardware-reduced support Samuel Ortiz
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 01/19] hw: i386: Decouple the ACPI build from the PC machine type Samuel Ortiz
2018-10-29 19:50 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 02/19] hw: acpi: Export ACPI build alignment API Samuel Ortiz
2018-10-29 19:01 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 03/19] hw: acpi: Export the RSDP build API Samuel Ortiz
2018-10-29 18:58 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 04/19] hw: acpi: Implement XSDT support for RSDP Samuel Ortiz
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 05/19] hw: arm: Switch to the AML build RSDP building routine Samuel Ortiz
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 06/19] hw: acpi: Generalize AML build routines Samuel Ortiz
2018-10-29 19:09 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 07/19] hw: acpi: Factorize _OSC AML across architectures Samuel Ortiz
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 08/19] hw: i386: Refactor PCI host getter Samuel Ortiz
2018-10-29 19:18 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 09/19] hw: acpi: Export and generalize the PCI host AML API Samuel Ortiz
2018-10-29 19:23 ` Philippe Mathieu-Daudé
2018-10-29 19:29 ` Philippe Mathieu-Daudé
2018-10-30 14:57 ` Samuel Ortiz
2018-10-30 18:04 ` Philippe Mathieu-Daudé
2018-10-31 10:46 ` Samuel Ortiz
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 10/19] hw: acpi: Export the MCFG getter Samuel Ortiz
2018-10-29 19:31 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 11/19] hw: acpi: Do not create hotplug method when handler is not defined Samuel Ortiz
2018-10-29 19:33 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 12/19] hw: i386: Make the hotpluggable memory size property more generic Samuel Ortiz
2018-10-29 19:38 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 13/19] hw: acpi: Export the SRAT AML build API Samuel Ortiz
2018-10-29 19:40 ` Philippe Mathieu-Daudé [this message]
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 14/19] hw: acpi: Fix memory hotplug AML generation error Samuel Ortiz
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 15/19] hw: acpi: Export the PCI hotplug API Samuel Ortiz
2018-10-29 19:46 ` Philippe Mathieu-Daudé
2018-10-29 23:34 ` Boeuf, Sebastien
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 16/19] hw: acpi: Retrieve the PCI bus from AcpiPciHpState Samuel Ortiz
2018-10-29 21:02 ` Philippe Mathieu-Daudé
2018-10-29 23:35 ` Boeuf, Sebastien
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 17/19] hw: acpi: Define ACPI tables builder interface Samuel Ortiz
2018-10-29 21:06 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 18/19] hw: i386: Export the MADT build method Samuel Ortiz
2018-10-29 21:18 ` Philippe Mathieu-Daudé
2018-10-29 17:01 ` [Qemu-devel] [PATCH v3 19/19] hw: i386: Implement the ACPI builder interface for PC Samuel Ortiz
2018-10-29 17:28 ` Paolo Bonzini
2018-10-30 14:13 ` Samuel Ortiz
2018-10-30 16:03 ` Paolo Bonzini
2018-10-31 1:02 ` Samuel Ortiz
2018-10-29 21:23 ` 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=e7cc3caf-3ee6-c9a8-0c2d-482f06a02019@redhat.com \
--to=philmd@redhat.com \
--cc=ehabkost@redhat.com \
--cc=imammedo@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-arm@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=rth@twiddle.net \
--cc=sameo@linux.intel.com \
--cc=yang.zhong@intel.com \
--cc=zhaoshenglong@huawei.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).