memory-device: Some cleanups
cover memory-device: Some cleanups
Working on adding multi-memslot support for virtio-mem (teaching memory device code about memory devices that can consume multiple memslots), I have some preparatory cleanups in my queue that make sense independent of the actual memory-device/virtio-mem extensions.
v1 -> v2:
- Allocate ms->device_memory only if the size > 0.
- Split it up and include more cleanups
David Hildenbrand (10):
memory-device: Unify enabled vs. supported error messages
memory-device: Introduce memory_devices_init() hw/arm/virt: Use memory_devices_init()
hw/ppc/spapr: Use memory_devices_init()
hw/loongarch/virt: Use memory_devices_init()
hw/i386/pc: Use memory_devices_init()
hw/i386/acpi-build: Rely on machine->device_memory when building SRAT hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
memory-device: Refactor memory_device_pre_plug()
memory-device: Track used region size in DeviceMemoryState
hw/arm/virt.c | 9 +---- hw/i386/acpi-build.c | 9 ++--- hw/i386/pc.c | 36 +++--------------- hw/loongarch/virt.c | 14 ++----- hw/mem/memory-device.c | 69 +++++++++++++++------------------- hw/ppc/spapr.c | 37 +++++++++--------- hw/ppc/spapr_hcall.c | 2 +- include/hw/boards.h | 2 + include/hw/i386/pc.h | 1 - include/hw/mem/memory-device.h | 2 + 10 files changed, 68 insertions(+), 113 deletions(-)
Signature · 1 line
2.40.1
PATCH 1/10 memory-device: Unify enabled vs. supported error messages
Let's unify the error messages, such that we can simply stop allocating ms->device_memory if the size would be 0 (and there are no memory devices ever).
The case of "not supported by the machine" should barely pop up either
way: if the machine doesn't support memory devices, it usually doesn't call the pre_plug handler ...
--- hw/mem/memory-device.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-)
hw/mem/memory-device.c +4 -9
@@ -104,15 +104,10 @@ static uint64_t memory_device_get_free_addr(MachineState *ms,GSList *list =NULL , *item;Range as, new = range_empty; -if (!ms->device_memory) { -error_setg (errp,"memory devices (e.g. for memory hotplug) are not " -"supported by the machine" ); -return 0 ; - } - -if (!memory_region_size (&ms->device_memory->mr)) { -error_setg (errp,"memory devices (e.g. for memory hotplug) are not " -"enabled, please specify the maxmem option" ); +if (!ms->device_memory || !memory_region_size (&ms->device_memory->mr)) { +error_setg (errp,"the configuration is not prepared for memory devices" +" (e.g., for memory hotplug), consider specifying the" +" maxmem option" );return 0 ; } range_init_nofail(&as, ms->device_memory->base,
Signature · 1 line
2.40.1
Re: [PATCH 01/10] memory-device: Unify enabled vs. supported error messages
Quoted 4 lines · David Hildenbrand
Let's unify the error messages, such that we can simply stop allocating
ms->device_memoryif the size would be 0 (and there are no memory devices ever).
Quoted 8 lines
The case of "not supported by the machine" should barely pop up either
way: if the machine doesn't support memory devices, it usually doesn't call thepre_plughandler ...Signed-off-by David Hildenbrand <[email protected]>--- hw/mem/memory-device.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-)
PATCH 2/10 memory-device: Introduce memory_devices_init()
In reply to David HildenbrandLet's intrduce a new helper that we will use to replace existing memory device setup code during machine initialization. We'll enforce that the size has to be > 0.
Once all machines were converted, we'll only allocate ms->device_memory if the size > 0.
--- hw/mem/memory-device.c | 14 ++++++++++++++ include/hw/mem/memory-device.h | 2 ++ 2 files changed, 16 insertions(+)
hw/mem/memory-device.c +14 -0
@@ -17,6 +17,7 @@#include "qemu/range.h" #include "hw/virtio/vhost.h" #include "sysemu/kvm.h" +#include "exec/address-spaces.h" #include "trace.h" static gint memory_device_addr_sort (gconstpointer a,gconstpointer b)
@@ -328,6 +329,19 @@ uint64_t memory_device_get_region_size(const MemoryDeviceState *md,return memory_region_size (mr); } +void memory_devices_init (MachineState *ms,hwaddr base,uint64_t size) +{ +g_assert (size); +g_assert (!ms->device_memory); + ms->device_memory =g_new0 (DeviceMemoryState,1 ); + ms->device_memory->base = base; + +memory_region_init (&ms->device_memory->mr,OBJECT (ms),"device-memory" , + size); +memory_region_add_subregion (get_system_memory (), ms->device_memory->base, + &ms->device_memory->mr); +} +static const TypeInfo memory_device_info = { .name =TYPE_MEMORY_DEVICE , .parent =TYPE_INTERFACE ,
include/hw/mem/memory-device.h +2 -0
@@ -16,6 +16,7 @@#include "hw/qdev-core.h" #include "qapi/qapi-types-machine.h" #include "qom/object.h" +#include "exec/hwaddr.h" #define TYPE_MEMORY_DEVICE "memory-device"
@@ -113,5 +114,6 @@ void memory_device_plug(MemoryDeviceState *md, MachineState *ms);void memory_device_unplug (MemoryDeviceState *md,MachineState *ms);uint64_t memory_device_get_region_size (const MemoryDeviceState *md,Error **errp); +void memory_devices_init (MachineState *ms,hwaddr base,uint64_t size);#endif
Signature · 1 line
2.40.1
Re: [PATCH 02/10] memory-device: Introduce memory_devices_init()
Quoted 4 lines · David Hildenbrand
Let's intrduce a new helper that we will use to replace existing memory device setup code during machine initialization. We'll enforce that the size has to be > 0.
Quoted 8 lines
Once all machines were converted, we'll only allocate
ms->device_memoryif the size > 0.Signed-off-by David Hildenbrand <[email protected]>--- hw/mem/memory-device.c | 14 ++++++++++++++ include/hw/mem/memory-device.h | 2 ++ 2 files changed, 16 insertions(+)
Re: [PATCH 02/10] memory-device: Introduce memory_devices_init()
In reply to David HildenbrandHi David,
Quoted 27 lines · David Hildenbrand
Let's intrduce a new helper that we will use to replace existing memory device setup code during machine initialization. We'll enforce that the size has to be > 0.
Once all machines were converted, we'll only allocate
ms->device_memoryif the size > 0.Signed-off-by David Hildenbrand <[email protected]>--- hw/mem/memory-device.c | 14 ++++++++++++++ include/hw/mem/memory-device.h | 2 ++ 2 files changed, 16 insertions(+)include/hw/mem/memory-device.h +1
-0@@ -16,6 +16,7 @@#include "hw/qdev-core.h" #include "qapi/qapi-types-machine.h" #include "qom/object.h" +#include "exec/hwaddr.h" #define TYPE_MEMORY_DEVICE "memory-device"
include/hw/mem/memory-device.h void memory_device_plug(MemoryDeviceState *md, MachineState *ms); · line 117
Quoted 4 lines
include/hw/mem/memory-device.h +1
-0@@ -113,5 +114,6 @@ void memory_device_plug(MemoryDeviceState *md, MachineState *ms);void memory_device_unplug (MemoryDeviceState *md,MachineState *ms);uint64_t memory_device_get_region_size (const MemoryDeviceState *md,Error **errp); +void memory_devices_init (MachineState *ms,hwaddr base,uint64_t size);
While hw/mem/memory-device.c contains the implementation, all callers are expected to be around Machine object, right? Thus maybe this _init() could be declared in "hw/boards.h", already included by machines (eventually renaming as machine_init_memory_devices() ). Then machines implementation don't have to all include "hw/mem/memory-device.h".
Alternatively, keep memory_devices_init() declared here, but implement machine_init_memory_devices() in hw/core/machine.c, declaring it in "hw/boards.h", so again machines don't have to include "hw/mem/memory-device.h".
What do you think?
Re: [PATCH 02/10] memory-device: Introduce memory_devices_init()
Quoted 29 lines · Philippe Mathieu-Daudé
Hi David,
Quoted 26 lines · David Hildenbrand
Let's intrduce a new helper that we will use to replace existing memory device setup code during machine initialization. We'll enforce that the size has to be > 0.
Once all machines were converted, we'll only allocate
ms->device_memoryif the size > 0.Signed-off-by David Hildenbrand <[email protected]>--- hw/mem/memory-device.c | 14 ++++++++++++++ include/hw/mem/memory-device.h | 2 ++ 2 files changed, 16 insertions(+)include/hw/mem/memory-device.h +1
-0@@ -16,6 +16,7 @@#include "hw/qdev-core.h" #include "qapi/qapi-types-machine.h" #include "qom/object.h" +#include "exec/hwaddr.h" #define TYPE_MEMORY_DEVICE "memory-device"
Quoted 12 lines
include/hw/mem/memory-device.h void memory_device_plug(MemoryDeviceState *md, MachineState *ms); · line 117
Quoted 5 lines
include/hw/mem/memory-device.h +1
-0@@ -113,5 +114,6 @@ void memory_device_plug(MemoryDeviceState *md, MachineState *ms);void memory_device_unplug (MemoryDeviceState *md,MachineState *ms);uint64_t memory_device_get_region_size (const MemoryDeviceState *md,Error **errp); +void memory_devices_init (MachineState *ms,hwaddr base,uint64_t size);While
hw/mem/memory-device.ccontains the implementation, all callers are expected to be around Machine object, right? Thus maybe this_init()could be declared in "hw/boards.h", already included by machines (eventually renaming asmachine_init_memory_devices()). Then machines implementation don't have to all include "hw/mem/memory-device.h".
Some (arm, i386) want to call the hotplug handle functions either way, so they'll still have to include that header.
But sure, we can rename to machine_init_memory_devices() and declare it include/hw/boards.h!
Quoted 5 lines
Alternatively, keep
memory_devices_init()declared here, but implementmachine_init_memory_devices()inhw/core/machine.c, declaring it in "hw/boards.h", so again machines don't have to include "hw/mem/memory-device.h".
I prefer the implementation residing in hw/mem/memory-device.c. I have some upcoming patches that add more "meat" to the init function, such that makes it looks cleaner to me to just reside hw/mem/memory-device.h.
Thanks!
Signature · 3 lines
Thanks, David / dhildenb
Re: [PATCH 02/10] memory-device: Introduce memory_devices_init()
Quoted 43 lines · David Hildenbrand
Quoted 29 lines · Philippe Mathieu-Daudé
Hi David,
Quoted 26 lines · David Hildenbrand
Let's intrduce a new helper that we will use to replace existing memory device setup code during machine initialization. We'll enforce that the size has to be > 0.
Once all machines were converted, we'll only allocate
ms->device_memoryif the size > 0.Signed-off-by David Hildenbrand <[email protected]>--- hw/mem/memory-device.c | 14 ++++++++++++++ include/hw/mem/memory-device.h | 2 ++ 2 files changed, 16 insertions(+)include/hw/mem/memory-device.h +1
-0@@ -16,6 +16,7 @@#include "hw/qdev-core.h" #include "qapi/qapi-types-machine.h" #include "qom/object.h" +#include "exec/hwaddr.h" #define TYPE_MEMORY_DEVICE "memory-device"Quoted 12 lines
include/hw/mem/memory-device.h void memory_device_plug(MemoryDeviceState *md, MachineState *ms); · line 117
Quoted 5 lines
include/hw/mem/memory-device.h +1
-0@@ -113,5 +114,6 @@ void memory_device_plug(MemoryDeviceState *md, MachineState *ms);void memory_device_unplug (MemoryDeviceState *md,MachineState *ms);uint64_t memory_device_get_region_size (const MemoryDeviceState *md,Error **errp); +void memory_devices_init (MachineState *ms,hwaddr base,uint64_t size);While
hw/mem/memory-device.ccontains the implementation, all callers are expected to be around Machine object, right? Thus maybe this_init()could be declared in "hw/boards.h", already included by machines (eventually renaming asmachine_init_memory_devices()). Then machines implementation don't have to all include "hw/mem/memory-device.h".
Quoted 5 lines
Some (arm, i386) want to call the hotplug handle functions either way, so they'll still have to include that header.
But sure, we can rename to
machine_init_memory_devices()and declare itinclude/hw/boards.h!
FWIW, I went with "machine_memory_devices_init()", to mach the style of "machine_run_board_init()".
Signature · 3 lines
Thanks, David / dhildenb
PATCH 3/10 hw/arm/virt: Use memory_devices_init()
In reply to David HildenbrandLet's use our new helper. We'll add the subregion to system RAM now earlier. That shouldn't matter, because the system RAM memory region should already be alive at that point.
--- hw/arm/virt.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-)
hw/arm/virt.c +1 -8
@@ -1817,10 +1817,7 @@ static void virt_set_memmap(VirtMachineState *vms, int pa_bits)virt_set_high_memmap (vms, base, pa_bits);if (device_memory_size >0 ) { - ms->device_memory =g_malloc0 (sizeof (*ms->device_memory)); - ms->device_memory->base = device_memory_base; -memory_region_init (&ms->device_memory->mr,OBJECT (vms), -"device-memory" , device_memory_size); +memory_devices_init (ms, device_memory_base, device_memory_size); } }
@@ -2261,10 +2258,6 @@ static void machvirt_init(MachineState *machine)memory_region_add_subregion (sysmem, vms->memmap[VIRT_MEM ].base, machine->ram); -if (machine->device_memory) { -memory_region_add_subregion (sysmem, machine->device_memory->base, - &machine->device_memory->mr); - }virt_flash_fdt (vms, sysmem, secure_sysmem ?: sysmem);
Signature · 1 line
2.40.1
Re: [PATCH 03/10] hw/arm/virt: Use memory_devices_init()
Quoted 9 lines · David Hildenbrand
Let's use our new helper. We'll add the subregion to system RAM now earlier. That shouldn't matter, because the system RAM memory region should already be alive at that point.
Cc Peter Maydell <[email protected]>Signed-off-by David Hildenbrand <[email protected]>--- hw/arm/virt.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-)
PATCH 4/10 hw/ppc/spapr: Use memory_devices_init()
In reply to David HildenbrandLet's use our new helper and stop always allocating ms->device_memory. There is no difference in common memory-device code anymore between ms->device_memory being NULL or the size being 0. So we only have to teach spapr code that ms->device_memory isn't always around.
We can now modify two maxram_size checks to rely on ms->device_memory for detecting whether we have memory devices.
--- hw/ppc/spapr.c | 37 +++++++++++++++++++------------------ hw/ppc/spapr_hcall.c | 2 +- 2 files changed, 20 insertions(+), 19 deletions(-)
hw/ppc/spapr.c +19 -18
@@ -546,10 +546,8 @@ static int spapr_dt_dynamic_reconfiguration_memory(SpaprMachineState *spapr, cpu_to_be32(lmb_size & 0xffffffff)};MemoryDeviceInfoList *dimms =NULL ; -/* -* Don't create the node if there is no device memory -*/ -if (machine->ram_size == machine->maxram_size) { +/* Don't create the node if there is no device memory. */ +if (!machine->device_memory) {return 0 ; }
@@ -859,16 +857,23 @@ static void spapr_dt_rtas(SpaprMachineState *spapr, void *fdt)int rtas;GString *hypertas =g_string_sized_new (256 );GString *qemu_hypertas =g_string_sized_new (256 ); -uint64_t max_device_addr =MACHINE (spapr)->device_memory->base + -memory_region_size (&MACHINE (spapr)->device_memory->mr);uint32_t lrdr_capacity[] = { -cpu_to_be32 (max_device_addr >>32 ), -cpu_to_be32 (max_device_addr &0xffffffff ), +0 , +0 ,cpu_to_be32 (SPAPR_MEMORY_BLOCK_SIZE >>32 ),cpu_to_be32 (SPAPR_MEMORY_BLOCK_SIZE &0xffffffff ),cpu_to_be32 (ms->smp.max_cpus / ms->smp.threads), }; +/* Do we have device memory? */ +if (MACHINE (spapr)->device_memory) { +uint64_t max_device_addr =MACHINE (spapr)->device_memory->base + +memory_region_size (&MACHINE (spapr)->device_memory->mr); + + lrdr_capacity[0 ] =cpu_to_be32 (max_device_addr >>32 ); + lrdr_capacity[1 ] =cpu_to_be32 (max_device_addr &0xffffffff ); + } +_FDT (rtas =fdt_add_subnode (fdt,0 ,"rtas" ));/* hypertas */
@@ -2454,6 +2459,7 @@ static void spapr_create_lmb_dr_connectors(SpaprMachineState *spapr)uint32_t nr_lmbs = (machine->maxram_size - machine->ram_size)/lmb_size;int i; +g_assert (!nr_lmbs || machine->device_memory);for (i =0 ; i < nr_lmbs; i++) {uint64_t addr;
@@ -2866,12 +2872,11 @@ static void spapr_machine_init(MachineState *machine)/* map RAM */ memory_region_add_subregion (sysmem,0 , machine->ram); -/* always allocate the device memory information */ - machine->device_memory =g_malloc0 (sizeof (*machine->device_memory)); -/* initialize hotplug memory address space */ if (machine->ram_size < machine->maxram_size) {ram_addr_t device_mem_size = machine->maxram_size - machine->ram_size; +hwaddr device_mem_base; +/* * Limit the number of hotpluggable memory slots to half the number * slots that KVM supports, leaving the other half for PCI and other
@@ -2890,12 +2895,8 @@ static void spapr_machine_init(MachineState *machine)exit (1 ); } - machine->device_memory->base =ROUND_UP (machine->ram_size, -SPAPR_DEVICE_MEM_ALIGN ); -memory_region_init (&machine->device_memory->mr,OBJECT (spapr), -"device-memory" , device_mem_size); -memory_region_add_subregion (sysmem, machine->device_memory->base, - &machine->device_memory->mr); + device_mem_base =ROUND_UP (machine->ram_size,SPAPR_DEVICE_MEM_ALIGN ); +memory_devices_init (machine, device_mem_base, device_mem_size); }if (smc->dr_lmb_enabled) {
@@ -5109,7 +5110,7 @@ static bool phb_placement_2_7(SpaprMachineState *spapr, uint32_t index,int i;/* Do we have device memory? */ -if (MACHINE (spapr)->maxram_size > ram_top) { +if (MACHINE (spapr)->device_memory) {/* Can't just use maxram_size, because there may be an * alignment gap between normal and device memory regions */
hw/ppc/spapr_hcall.c +1 -1
@@ -31,7 +31,7 @@ bool is_ram_address(SpaprMachineState *spapr, hwaddr addr)if (addr < machine->ram_size) {return true; } -if ((addr >= dms->base) +if (dms && (addr >= dms->base) && ((addr - dms->base) <memory_region_size (&dms->mr))) {return true; }
Signature · 1 line
2.40.1
Re: [PATCH 04/10] hw/ppc/spapr: Use memory_devices_init()
Quoted 14 lines · David Hildenbrand
Let's use our new helper and stop always allocating
ms->device_memory. There is no difference in common memory-device code anymore betweenms->device_memorybeing NULL or the size being 0. So we only have to teach spapr code thatms->device_memoryisn't always around.We can now modify two
maxram_sizechecks to rely onms->device_memoryfor detecting whether we have memory devices.Cc Daniel Henrique Barboza <[email protected]>Cc "Cédric Le Goater" <[email protected]>Cc David Gibson <[email protected]>Cc Greg Kurz <[email protected]>Cc Harsh Prateek Bora <[email protected]>Signed-off-by David Hildenbrand <[email protected]>
Quoted 4 lines
--- hw/ppc/spapr.c | 37 +++++++++++++++++++------------------ hw/ppc/spapr_hcall.c | 2 +- 2 files changed, 20 insertions(+), 19 deletions(-)
PATCH 5/10 hw/loongarch/virt: Use memory_devices_init()
In reply to David HildenbrandLet's use our new helper.
--- hw/loongarch/virt.c | 14 ++++---------- 1 file changed, 4 insertions(+), 10 deletions(-)
hw/loongarch/virt.c +4 -10
@@ -45,7 +45,7 @@#include "sysemu/block-backend.h" #include "hw/block/flash.h" #include "qemu/error-report.h" - +#include "hw/mem/memory-device.h" static void virt_flash_create (LoongArchMachineState *lams) {
@@ -805,8 +805,8 @@ static void loongarch_init(MachineState *machine)/* initialize device memory address space */ if (machine->ram_size < machine->maxram_size) { - machine->device_memory =g_malloc0 (sizeof (*machine->device_memory));ram_addr_t device_mem_size = machine->maxram_size - machine->ram_size; +hwaddr device_mem_base;if (machine->ram_slots >ACPI_MAX_RAM_SLOTS ) { error_report("unsupported amount of memory slots: %"PRIu64,
@@ -821,14 +821,8 @@ static void loongarch_init(MachineState *machine)exit (EXIT_FAILURE ); }/* device memory base is the top of high memory address. */ - machine->device_memory->base =0x90000000 + highram_size; - machine->device_memory->base = -ROUND_UP (machine->device_memory->base,1 * GiB); - -memory_region_init (&machine->device_memory->mr,OBJECT (lams), -"device-memory" , device_mem_size); -memory_region_add_subregion (address_space_mem, machine->device_memory->base, - &machine->device_memory->mr); + device_mem_base =ROUND_UP (0x90000000 + highram_size,1 * GiB); +memory_devices_init (machine, device_mem_base, device_mem_size); }/* Add isa io region */
Signature · 1 line
2.40.1
Re: [PATCH 05/10] hw/loongarch/virt: Use memory_devices_init()
Quoted 8 lines · David Hildenbrand
Let's use our new helper.
Cc Xiaojuan Yang <[email protected]>Cc Song Gao <[email protected]>Signed-off-by David Hildenbrand <[email protected]>--- hw/loongarch/virt.c | 14 ++++---------- 1 file changed, 4 insertions(+), 10 deletions(-)
Re: [PATCH 05/10] hw/loongarch/virt: Use memory_devices_init()
In reply to David HildenbrandQuoted 35 lines · David Hildenbrand
Let's use our new helper.
Cc Xiaojuan Yang <[email protected]>Cc Song Gao <[email protected]>Signed-off-by David Hildenbrand <[email protected]>--- hw/loongarch/virt.c | 14 ++++---------- 1 file changed, 4 insertions(+), 10 deletions(-)hw/loongarch/virt.c +2
-2@@ -45,7 +45,7 @@#include "sysemu/block-backend.h" #include "hw/block/flash.h" #include "qemu/error-report.h" - +#include "hw/mem/memory-device.h" static void virt_flash_create (LoongArchMachineState *lams) {@@ -805,8 +805,8 @@ static void loongarch_init(MachineState *machine)/* initialize device memory address space */ if (machine->ram_size < machine->maxram_size) { - machine->device_memory =g_malloc0 (sizeof (*machine->device_memory));ram_addr_t device_mem_size = machine->maxram_size - machine->ram_size; +hwaddr device_mem_base;if (machine->ram_slots >ACPI_MAX_RAM_SLOTS ) { error_report("unsupported amount of memory slots: %"PRIu64,@@ -821,14 +821,8 @@ static void loongarch_init(MachineState *machine)exit (EXIT_FAILURE ); }
hw/loongarch/virt.c static void loongarch_init(MachineState *machine) · line 824
Quoted 10 lines
hw/loongarch/virt.c +1
-8/* device memory base is the top of high memory address. */ - machine->device_memory->base =0x90000000 + highram_size; - machine->device_memory->base = -ROUND_UP (machine->device_memory->base,1 * GiB); - -memory_region_init (&machine->device_memory->mr,OBJECT (lams), -"device-memory" , device_mem_size); -memory_region_add_subregion (address_space_mem, machine->device_memory->base, - &machine->device_memory->mr); + device_mem_base =ROUND_UP (0x90000000 + highram_size,1 * GiB);
Use VIRT_HIGHMEM_BASE.
Otherwise,
Thanks.
Song Gao
Quoted 4 lines
hw/loongarch/virt.c +1
-0+memory_devices_init (machine, device_mem_base, device_mem_size); }/* Add isa io region */
PATCH 6/10 hw/i386/pc: Use memory_devices_init()
In reply to David HildenbrandLet's use our new helper and stop always allocating ms->device_memory. Once allcoated, we're sure that the size > 0 and that the base was initialized.
Adjust the code in pc_memory_init() to check for machine->device_memory instead of pcmc->has_reserved_memory and machine->device_memory->base.
--- hw/i386/pc.c | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-)
hw/i386/pc.c +5 -12
@@ -1039,13 +1039,11 @@ void pc_memory_init(PCMachineState *pcms,exit (EXIT_FAILURE ); } -/* always allocate the device memory information */ - machine->device_memory =g_malloc0 (sizeof (*machine->device_memory)); -/* initialize device memory address space */ if (pcmc->has_reserved_memory && (machine->ram_size < machine->maxram_size)) {ram_addr_t device_mem_size; +hwaddr device_mem_base;if (machine->ram_slots >ACPI_MAX_RAM_SLOTS ) { error_report("unsupported amount of memory slots: %"PRIu64,
@@ -1060,19 +1058,14 @@ void pc_memory_init(PCMachineState *pcms,exit (EXIT_FAILURE ); } -pc_get_device_memory_range (pcms, &machine->device_memory->base, &device_mem_size); +pc_get_device_memory_range (pcms, &device_mem_base, &device_mem_size); -if ((machine->device_memory->base + device_mem_size) < - device_mem_size) { +if (device_mem_base + device_mem_size < device_mem_size) {error_report ("unsupported amount of maximum memory: " RAM_ADDR_FMT , machine->maxram_size);exit (EXIT_FAILURE ); } - -memory_region_init (&machine->device_memory->mr,OBJECT (pcms), -"device-memory" , device_mem_size); -memory_region_add_subregion (system_memory, machine->device_memory->base, - &machine->device_memory->mr); +memory_devices_init (machine, device_mem_base, device_mem_size); }if (pcms->cxl_devices_state.is_enabled) {
@@ -1120,7 +1113,7 @@ void pc_memory_init(PCMachineState *pcms,rom_set_fw (fw_cfg); -if (pcmc->has_reserved_memory && machine->device_memory->base) { +if (machine->device_memory) {uint64_t *val =g_malloc (sizeof (*val));PCMachineClass *pcmc =PC_MACHINE_GET_CLASS (pcms);uint64_t res_mem_end = machine->device_memory->base;
Signature · 1 line
2.40.1
Re: [PATCH 06/10] hw/i386/pc: Use memory_devices_init()
Quoted 13 lines · David Hildenbrand
Let's use our new helper and stop always allocating
ms->device_memory. Once allcoated, we're sure that the size > 0 and that the base was initialized.Adjust the code in
pc_memory_init()to check formachine->device_memoryinstead ofpcmc->has_reserved_memoryandmachine->device_memory->base.Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>
Quoted 3 lines
--- hw/i386/pc.c | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-)
PATCH 7/10 hw/i386/acpi-build: Rely on machine->device_memory when building SRAT
In reply to David HildenbrandWe're already looking at machine->device_memory when calling build_srat_memory(), so let's simply avoid going via PC_MACHINE_DEVMEM_REGION_SIZE to get the size and rely on machine->device_memory directly.
Once machine->device_memory is set, we know that the size > 0. The code now looks much more similar the hw/arm/virt-acpi-build.c variant.
--- hw/i386/acpi-build.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-)
hw/i386/acpi-build.c +3 -6
@@ -1950,12 +1950,8 @@ build_srat(GArray *table_data, BIOSLinker *linker, MachineState *machine)MachineClass *mc =MACHINE_GET_CLASS (machine);X86MachineState *x86ms =X86_MACHINE (machine);const CPUArchIdList *apic_ids = mc->possible_cpu_arch_ids (machine); -PCMachineState *pcms =PC_MACHINE (machine);int nb_numa_nodes = machine->numa_state->num_nodes;NodeInfo *numa_info = machine->numa_state->nodes; -ram_addr_t hotpluggable_address_space_size = -object_property_get_int (OBJECT (pcms),PC_MACHINE_DEVMEM_REGION_SIZE , -NULL );AcpiTable table = { .sig ="SRAT" , .rev =1 , .oem_id = x86ms->oem_id, .oem_table_id = x86ms->oem_table_id };
@@ -2071,9 +2067,10 @@ build_srat(GArray *table_data, BIOSLinker *linker, MachineState *machine)* Memory devices may override proximity set by this entry, * providing _PXM method if necessary. */ -if (hotpluggable_address_space_size) { +if (machine->device_memory) {build_srat_memory (table_data, machine->device_memory->base, - hotpluggable_address_space_size, nb_numa_nodes -1 , +memory_region_size (&machine->device_memory->mr), + nb_numa_nodes -1 ,MEM_AFFINITY_HOTPLUGGABLE |MEM_AFFINITY_ENABLED ); }
Signature · 1 line
2.40.1
Re: [PATCH 07/10] hw/i386/acpi-build: Rely on machine->device_memory when building SRAT
Quoted 14 lines · David Hildenbrand
We're already looking at
machine->device_memorywhen callingbuild_srat_memory(), so let's simply avoid going viaPC_MACHINE_DEVMEM_REGION_SIZEto get the size and rely onmachine->device_memorydirectly.Once
machine->device_memoryis set, we know that the size > 0. The code now looks much more similar thehw/arm/virt-acpi-build.cvariant.Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>
Quoted 3 lines
--- hw/i386/acpi-build.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-)
PATCH 8/10 hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
In reply to David HildenbrandThere are no remaining users in the tree, so let's remove it.
--- hw/i386/pc.c | 19 ------------------- include/hw/i386/pc.h | 1 - 2 files changed, 20 deletions(-)
hw/i386/pc.c +0 -19
@@ -1646,21 +1646,6 @@ static HotplugHandler *pc_get_hotplug_handler(MachineState *machine,return NULL ; } -static void -pc_machine_get_device_memory_region_size (Object *obj,Visitor *v, -const char *name,void *opaque, -Error **errp) -{ -MachineState *ms =MACHINE (obj); -int64_t value =0 ; - -if (ms->device_memory) { - value =memory_region_size (&ms->device_memory->mr); - } - -visit_type_int (v, name, &value, errp); -} -static void pc_machine_get_vmport (Object *obj,Visitor *v,const char *name,void *opaque,Error **errp) {
@@ -1980,10 +1965,6 @@ static void pc_machine_class_init(ObjectClass *oc, void *data)object_class_property_set_description (oc,PC_MACHINE_MAX_RAM_BELOW_4G ,"Maximum ram below the 4G boundary (32bit boundary)" ); -object_class_property_add (oc,PC_MACHINE_DEVMEM_REGION_SIZE ,"int" , - pc_machine_get_device_memory_region_size,NULL , -NULL ,NULL ); -object_class_property_add (oc,PC_MACHINE_VMPORT ,"OnOffAuto" , pc_machine_get_vmport, pc_machine_set_vmport,NULL ,NULL );
include/hw/i386/pc.h +0 -1
@@ -60,7 +60,6 @@ typedef struct PCMachineState {#define PC_MACHINE_ACPI_DEVICE_PROP "acpi-device"#define PC_MACHINE_MAX_RAM_BELOW_4G "max-ram-below-4g" -#define PC_MACHINE_DEVMEM_REGION_SIZE "device-memory-region-size"#define PC_MACHINE_VMPORT "vmport"#define PC_MACHINE_SMBUS "smbus"#define PC_MACHINE_SATA "sata"
Signature · 1 line
2.40.1
Re: [PATCH 08/10] hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
Quoted 8 lines · David Hildenbrand
There are no remaining users in the tree, so let's remove it.
Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>
Quoted 4 lines
--- hw/i386/pc.c | 19 ------------------- include/hw/i386/pc.h | 1 - 2 files changed, 20 deletions(-)
Re: [PATCH 08/10] hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
In reply to David HildenbrandQuoted 8 lines · David Hildenbrand
There are no remaining users in the tree, so let's remove it.
Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>
This (with previous patches) means any user changing device-memory-region-size machine property is now broken, right? How do we know there are no users?
Quoted 56 lines
--- hw/i386/pc.c | 19 ------------------- include/hw/i386/pc.h | 1 - 2 files changed, 20 deletions(-)hw/i386/pc.c +0
-19@@ -1646,21 +1646,6 @@ static HotplugHandler *pc_get_hotplug_handler(MachineState *machine,return NULL ; } -static void -pc_machine_get_device_memory_region_size (Object *obj,Visitor *v, -const char *name,void *opaque, -Error **errp) -{ -MachineState *ms =MACHINE (obj); -int64_t value =0 ; - -if (ms->device_memory) { - value =memory_region_size (&ms->device_memory->mr); - } - -visit_type_int (v, name, &value, errp); -} -static void pc_machine_get_vmport (Object *obj,Visitor *v,const char *name,void *opaque,Error **errp) {@@ -1980,10 +1965,6 @@ static void pc_machine_class_init(ObjectClass *oc, void *data)object_class_property_set_description (oc,PC_MACHINE_MAX_RAM_BELOW_4G ,"Maximum ram below the 4G boundary (32bit boundary)" ); -object_class_property_add (oc,PC_MACHINE_DEVMEM_REGION_SIZE ,"int" , - pc_machine_get_device_memory_region_size,NULL , -NULL ,NULL ); -object_class_property_add (oc,PC_MACHINE_VMPORT ,"OnOffAuto" , pc_machine_get_vmport, pc_machine_set_vmport,NULL ,NULL );include/hw/i386/pc.h +0
-1@@ -60,7 +60,6 @@ typedef struct PCMachineState {#define PC_MACHINE_ACPI_DEVICE_PROP "acpi-device"#define PC_MACHINE_MAX_RAM_BELOW_4G "max-ram-below-4g" -#define PC_MACHINE_DEVMEM_REGION_SIZE "device-memory-region-size"#define PC_MACHINE_VMPORT "vmport"#define PC_MACHINE_SMBUS "smbus"#define PC_MACHINE_SATA "sata"Signature · 1 line
2.40.1
Re: [PATCH 08/10] hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
Quoted 11 lines · Michael S. Tsirkin
Quoted 8 lines · David Hildenbrand
There are no remaining users in the tree, so let's remove it.
Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>
Quoted 2 lines
This (with previous patches) means any user changing device-memory-region-size machine property is now broken, right?
We only had a getter, no setter (for good reason).
Quoted 1 line
How do we know there are no users?
We don't. A quick google search makes "device-memory-region-size" and "qom-get" only pop up in BUG fixes for something that appears to be QEMU developer driven.
I don't consider it any useful, but if we want to be careful, sure we can leave it around.
Signature · 3 lines
Thanks, David / dhildenb
Re: [PATCH 08/10] hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
Quoted 17 lines · David Hildenbrand
Quoted 11 lines · Michael S. Tsirkin
Quoted 8 lines · David Hildenbrand
There are no remaining users in the tree, so let's remove it.
Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>Quoted 2 lines
This (with previous patches) means any user changing device-memory-region-size machine property is now broken, right?
We only had a getter, no setter (for good reason).
Quoted 8 lines
Quoted 1 line
How do we know there are no users?
We don't. A quick google search makes "device-memory-region-size" and "qom-get" only pop up in BUG fixes for something that appears to be QEMU developer driven.
I don't consider it any useful, but if we want to be careful, sure we can leave it around.
Looking further, libvirt doesn't use it (and never used it).
I already renamed it in 2018 without anybody complaining:
https://www.mail-archive.com/[email protected]/msg532101.html
So I'm quite confident that nobody will really miss this property.
Signature · 3 lines
Thanks, David / dhildenb
Re: [PATCH 08/10] hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
Quoted 27 lines · David Hildenbrand
Quoted 17 lines · David Hildenbrand
Quoted 11 lines · Michael S. Tsirkin
Quoted 8 lines · David Hildenbrand
There are no remaining users in the tree, so let's remove it.
Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>Quoted 2 lines
This (with previous patches) means any user changing device-memory-region-size machine property is now broken, right?
We only had a getter, no setter (for good reason).
Quoted 8 lines
Quoted 1 line
How do we know there are no users?
We don't. A quick google search makes "device-memory-region-size" and "qom-get" only pop up in BUG fixes for something that appears to be QEMU developer driven.
I don't consider it any useful, but if we want to be careful, sure we can leave it around.
Quoted 12 lines
Looking further, libvirt doesn't use it (and never used it).
I already renamed it in 2018 without anybody complaining:
https://www.mail-archive.com/[email protected]/msg532101.html
So I'm quite confident that nobody will really miss this property.
Signature · 3 lines
Thanks, David / dhildenb
OK. In the future we need to be careful and use "x-" prefix for what we don't want to expose.
Signature · 1 line
MST
Re: [PATCH 08/10] hw/i386/pc: Remove PC_MACHINE_DEVMEM_REGION_SIZE
In reply to David HildenbrandQuoted 12 lines · David Hildenbrand
Quoted 11 lines · Michael S. Tsirkin
Quoted 8 lines · David Hildenbrand
There are no remaining users in the tree, so let's remove it.
Cc Paolo Bonzini <[email protected]>Cc Richard Henderson <[email protected]>Cc Eduardo Habkost <[email protected]>Cc "Michael S. Tsirkin" <[email protected]>Cc Marcel Apfelbaum <[email protected]>Signed-off-by David Hildenbrand <[email protected]>
Quoted 8 lines
Quoted 2 lines
This (with previous patches) means any user changing device-memory-region-size machine property is now broken, right?
We only had a getter, no setter (for good reason).
Quoted 1 line
How do we know there are no users?
We don't. A quick google search makes "device-memory-region-size" and "qom-get" only pop up in BUG fixes for something that appears to be QEMU developer driven.
This was my analysis.
Quoted 1 line
I don't consider it any useful, but if we want to be careful, sure we can leave it around.
If we want to keep it, we should move it to generic code IMHO, not PC machine. Otherwise the less unused code the better :)
PATCH 9/10 memory-device: Refactor memory_device_pre_plug()
In reply to David HildenbrandLet's move memory_device_check_addable() and basic checks out of memory_device_get_free_addr() directly into memory_device_pre_plug().
Separating basic checks from address assignment is cleaner and prepares for further changes.
As all memory device users now use memory_devices_init(), and that function enforces that the size is 0, we can drop the check for an empty region.
--- hw/mem/memory-device.c | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-)
hw/mem/memory-device.c +14 -14
@@ -69,9 +69,10 @@ static int memory_device_used_region_size(Object *obj, void *opaque)return 0 ; } -static void memory_device_check_addable (MachineState *ms,uint64_t size, +static void memory_device_check_addable (MachineState *ms,MemoryRegion *mr,Error **errp) { +const uint64_t size =memory_region_size (mr);uint64_t used_region_size =0 ;/* we will need a new memory slot for kvm and vhost */
@@ -101,16 +102,9 @@ static uint64_t memory_device_get_free_addr(MachineState *ms, uint64_t align, uint64_t size, Error **errp) { - Error *err = NULL; GSList *list = NULL, *item;Range as, new = range_empty; -if (!ms->device_memory || !memory_region_size (&ms->device_memory->mr)) { -error_setg (errp,"the configuration is not prepared for memory devices" -" (e.g., for memory hotplug), consider specifying the" -" maxmem option" ); -return 0 ; - }range_init_nofail (&as, ms->device_memory->base,memory_region_size (&ms->device_memory->mr));
@@ -122,12 +116,6 @@ static uint64_t memory_device_get_free_addr(MachineState *ms, align); } -memory_device_check_addable (ms, size, &err); -if (err) { -error_propagate (errp, err); -return 0 ; - } -if (hint && !QEMU_IS_ALIGNED (*hint, align)) {error_setg (errp,"address must be aligned to 0x%" PRIx64" bytes" , align);
@@ -251,11 +239,23 @@ void memory_device_pre_plug(MemoryDeviceState *md, MachineState *ms,uint64_t addr, align =0 ;MemoryRegion *mr; +if (!ms->device_memory) { +error_setg (errp,"the configuration is not prepared for memory devices" +" (e.g., for memory hotplug), consider specifying the" +" maxmem option" ); +return ; + } + mr = mdc->get_memory_region (md, &local_err);if (local_err) { goto out; } +memory_device_check_addable (ms, mr, &local_err); +if (local_err) { + goto out; + } +if (legacy_align) { align = *legacy_align; }else {
Signature · 1 line
2.40.1
PATCH 10/10 memory-device: Track used region size in DeviceMemoryState
In reply to David HildenbrandLet's avoid iterating over all devices and simply track it in the DeviceMemoryState.
--- hw/mem/memory-device.c | 22 +++------------------- include/hw/boards.h | 2 ++ 2 files changed, 5 insertions(+), 19 deletions(-)
hw/mem/memory-device.c +3 -19
@@ -52,28 +52,11 @@ static int memory_device_build_list(Object *obj, void *opaque)return 0 ; } -static int memory_device_used_region_size (Object *obj,void *opaque) -{ -uint64_t *size = opaque; - -if (object_dynamic_cast (obj,TYPE_MEMORY_DEVICE )) { -const DeviceState *dev =DEVICE (obj); -const MemoryDeviceState *md =MEMORY_DEVICE (obj); - -if (dev->realized) { - *size +=memory_device_get_region_size (md, &error_abort); - } - } - -object_child_foreach (obj, memory_device_used_region_size, opaque); -return 0 ; -} -static void memory_device_check_addable (MachineState *ms,MemoryRegion *mr,Error **errp) { +const uint64_t used_region_size = ms->device_memory->used_region_size;const uint64_t size =memory_region_size (mr); -uint64_t used_region_size =0 ;/* we will need a new memory slot for kvm and vhost */ if (kvm_enabled () && !kvm_has_free_slot (ms)) {
@@ -86,7 +69,6 @@ static void memory_device_check_addable(MachineState *ms, MemoryRegion *mr, }/* will we exceed the total amount of memory specified */ -memory_device_used_region_size (OBJECT (ms), &used_region_size);if (used_region_size + size < used_region_size || used_region_size + size > ms->maxram_size - ms->ram_size) { error_setg(errp, "not enough space, currently 0x%" PRIx64
@@ -292,6 +274,7 @@ void memory_device_plug(MemoryDeviceState *md, MachineState *ms) mr = mdc->get_memory_region (md, &error_abort);g_assert (ms->device_memory); + ms->device_memory->used_region_size +=memory_region_size (mr);memory_region_add_subregion (&ms->device_memory->mr, addr - ms->device_memory->base, mr);trace_memory_device_plug (DEVICE (md)->id ?DEVICE (md)->id :"" , addr);
@@ -310,6 +293,7 @@ void memory_device_unplug(MemoryDeviceState *md, MachineState *ms)g_assert (ms->device_memory);memory_region_del_subregion (&ms->device_memory->mr, mr); + ms->device_memory->used_region_size -=memory_region_size (mr);trace_memory_device_unplug (DEVICE (md)->id ?DEVICE (md)->id :"" , mdc->get_addr (md)); }
include/hw/boards.h +2 -0
@@ -294,11 +294,13 @@ struct MachineClass {* address space for memory devices starts * @mr: address space container for memory devices * @dimm_size: the sum of plugged DIMMs' sizes +* @used_region_size: the part of @mr already used by memory devices */ typedef struct DeviceMemoryState {hwaddr base;MemoryRegion mr;uint64_t dimm_size; +uint64_t used_region_size; }DeviceMemoryState ;/**
Signature · 1 line
2.40.1
Re: [PATCH 10/10] memory-device: Track used region size in DeviceMemoryState
Quoted 8 lines · David Hildenbrand
Let's avoid iterating over all devices and simply track it in the DeviceMemoryState.
Signed-off-by David Hildenbrand <[email protected]>--- hw/mem/memory-device.c | 22 +++------------------- include/hw/boards.h | 2 ++ 2 files changed, 5 insertions(+), 19 deletions(-)
Nice.
This was supposed to carry the "v2" indicator ;)
Quoted 33 lines · David Hildenbrand
Working on adding multi-memslot support for virtio-mem (teaching memory device code about memory devices that can consume multiple memslots), I have some preparatory cleanups in my queue that make sense independent of the actual memory-device/virtio-mem extensions.
v1 -> v2:
- Allocatems->device_memoryonly if the size > 0.
- Split it up and include more cleanupsDavid Hildenbrand (10):
memory-device: Unify enabled vs. supported error messages
memory-device: Introducememory_devices_init()hw/arm/virt: Usememory_devices_init()
hw/ppc/spapr: Usememory_devices_init()hw/loongarch/virt: Usememory_devices_init()
hw/i386/pc: Usememory_devices_init()
hw/i386/acpi-build: Rely onmachine->device_memorywhen building SRAT hw/i386/pc: RemovePC_MACHINE_DEVMEM_REGION_SIZE
memory-device: Refactormemory_device_pre_plug()
memory-device: Track used region size in DeviceMemoryStatehw/arm/virt.c | 9 +---- hw/i386/acpi-build.c | 9 ++--- hw/i386/pc.c | 36 +++--------------- hw/loongarch/virt.c | 14 ++----- hw/mem/memory-device.c | 69 +++++++++++++++------------------- hw/ppc/spapr.c | 37 +++++++++--------- hw/ppc/spapr_hcall.c | 2 +- include/hw/boards.h | 2 + include/hw/i386/pc.h | 1 - include/hw/mem/memory-device.h | 2 + 10 files changed, 68 insertions(+), 113 deletions(-)
Signature · 3 lines
Thanks, David / dhildenb
Quoted 4 lines · David Hildenbrand
Working on adding multi-memslot support for virtio-mem (teaching memory device code about memory devices that can consume multiple memslots), I have some preparatory cleanups in my queue that make sense independent of the actual memory-device/virtio-mem extensions.
pc/acpi things:
Quoted 30 lines
v1 -> v2:
- Allocatems->device_memoryonly if the size > 0.
- Split it up and include more cleanupsDavid Hildenbrand (10):
memory-device: Unify enabled vs. supported error messages
memory-device: Introducememory_devices_init()hw/arm/virt: Usememory_devices_init()
hw/ppc/spapr: Usememory_devices_init()
hw/loongarch/virt: Usememory_devices_init()
hw/i386/pc: Usememory_devices_init()
hw/i386/acpi-build: Rely onmachine->device_memorywhen building SRAT hw/i386/pc: RemovePC_MACHINE_DEVMEM_REGION_SIZE
memory-device: Refactormemory_device_pre_plug()
memory-device: Track used region size in DeviceMemoryStatehw/arm/virt.c | 9 +---- hw/i386/acpi-build.c | 9 ++--- hw/i386/pc.c | 36 +++--------------- hw/loongarch/virt.c | 14 ++----- hw/mem/memory-device.c | 69 +++++++++++++++------------------- hw/ppc/spapr.c | 37 +++++++++--------- hw/ppc/spapr_hcall.c | 2 +- include/hw/boards.h | 2 + include/hw/i386/pc.h | 1 - include/hw/mem/memory-device.h | 2 + 10 files changed, 68 insertions(+), 113 deletions(-)Signature · 1 line
2.40.1
Quoted 9 lines · Michael S. Tsirkin
Quoted 4 lines · David Hildenbrand
Working on adding multi-memslot support for virtio-mem (teaching memory device code about memory devices that can consume multiple memslots), I have some preparatory cleanups in my queue that make sense independent of the actual memory-device/virtio-mem extensions.
pc/acpi things:
Thanks, I added that to patch 6/7/8.
Signature · 3 lines
Cheers, David / dhildenb