pc: inform SeaBIOS where 64-bit PCI hole begins
v2 cover pc: inform SeaBIOS where 64-bit PCI hole begins
* simplify PCI address space mapping into system address space, replacing code duplication in piix/q53 PCs with a helper function
* add fw_cfg 'etc/pcimem64-minimum-address' to allow QEMU reserve additional address space before 64-bit PCI hole. Which will be need for reserving memory hotplug region in highmem. SeaBIOS counterpart: http://patchwork.ozlabs.org/patch/283623/
v2:
* use negative priority to map PCI address space under RAM memory regions which allows simplify code by removing pci_hole & pci_hole64 memory region aliases
Series depends on:
"memory: Change MemoryRegion priorities from unsigned to signed:
Git tree for testing:
https://github.com/imammedo/qemu/commits/pcimem64-minimum-address-v2
Igor Mammedov (1):
pc: add 'etc/pcimem64-minimum-address' fw_cfg interface to SeaBIOS
Michael S. Tsirkin (1):
pc: map PCI address space as catchall region for not mapped addresses
hw/i386/pc.c | 28 ++++++++++++++++------------ hw/i386/pc_piix.c | 2 -- hw/pci-host/piix.c | 27 +++++---------------------- hw/pci-host/q35.c | 28 ++++++---------------------- include/hw/i386/pc.h | 15 +++------------ include/hw/pci-host/q35.h | 2 -- 6 files changed, 30 insertions(+), 72 deletions(-)
PATCH v2 1/2 pc: map PCI address space as catchall region for not mapped addresses
From: Michael S. Tsirkin <[email protected]>
With a help of negative memory region priority PCI address space is mapped underneath RAM regions effectively catching every access to addresses not mapped by any other region.
It simplifies PCI address space mapping into system address space.
--- hw/i386/pc.c | 20 ++++++-------------- hw/i386/pc_piix.c | 2 -- hw/pci-host/piix.c | 26 ++++---------------------- hw/pci-host/q35.c | 27 +++++---------------------- include/hw/i386/pc.h | 14 ++------------ include/hw/pci-host/q35.h | 2 -- 6 files changed, 17 insertions(+), 74 deletions(-)
hw/i386/pc.c +6 -14
@@ -1053,21 +1053,13 @@ PcGuestInfo *pc_guest_info_init(ram_addr_t below_4g_mem_size,return guest_info; } -void pc_init_pci64_hole (PcPciInfo *pci_info,uint64_t pci_hole64_start, -uint64_t pci_hole64_size) +/* setup pci memory address space mapping into system address space */ +void pc_pci_as_mapping_init (Object *owner,MemoryRegion *system_memory, +MemoryRegion *pci_address_space) { -if ((sizeof (hwaddr) ==4 ) || (!pci_hole64_size)) { -return ; - } -/* -* BIOS does not set MTRR entries for the 64 bit window, so no need to -* align address to power of two. Align address at 1G, this makes sure -* it can be exactly covered with a PAT entry even when using huge -* pages. -*/ - pci_info->w64.begin =ROUND_UP (pci_hole64_start,0x1ULL <<30 ); - pci_info->w64.end = pci_info->w64.begin + pci_hole64_size; -assert (pci_info->w64.begin <= pci_info->w64.end); +/* Set to lower priority than RAM */ +memory_region_add_subregion_overlap (system_memory,0x0 , + pci_address_space,-1 ); }void pc_acpi_init (const char *default_dsdt)
hw/i386/pc_piix.c +0 -2
@@ -146,8 +146,6 @@ static void pc_init1(QEMUMachineInitArgs *args,if (pci_enabled) { pci_bus =i440fx_init (&i440fx_state, &piix3_devfn, &isa_bus, gsi, system_memory, system_io, args->ram_size, - below_4g_mem_size, -0x100000000ULL - below_4g_mem_size, above_4g_mem_size, pci_memory, ram_memory); }else {
hw/pci-host/piix.c +4 -22
@@ -102,8 +102,6 @@ struct PCII440FXState {MemoryRegion *system_memory;MemoryRegion *pci_address_space;MemoryRegion *ram_memory; -MemoryRegion pci_hole; -MemoryRegion pci_hole_64bit;PAMMemoryRegion pam_regions[13 ];MemoryRegion smram_region;uint8_t smm_enabled;
@@ -312,8 +310,6 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state, MemoryRegion *address_space_mem, MemoryRegion *address_space_io, ram_addr_t ram_size, - hwaddr pci_hole_start, - hwaddr pci_hole_size, ram_addr_t above_4g_mem_size, MemoryRegion *pci_address_space, MemoryRegion *ram_memory)
@@ -326,7 +322,6 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state,PCII440FXState *f;unsigned i;I440FXState *i440fx; -uint64_t pci_hole64_size; dev =qdev_create (NULL ,TYPE_I440FX_PCI_HOST_BRIDGE ); s =PCI_HOST_BRIDGE (dev);
@@ -354,23 +349,10 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state, i440fx->pci_info.w32.begin =0xe0000000 ; } -memory_region_init_alias (&f->pci_hole,OBJECT (d),"pci-hole" , f->pci_address_space, - pci_hole_start, pci_hole_size); -memory_region_add_subregion (f->system_memory, pci_hole_start, &f->pci_hole); - - pci_hole64_size =pci_host_get_hole64_size (i440fx->pci_hole64_size); - -pc_init_pci64_hole (&i440fx->pci_info,0x100000000ULL + above_4g_mem_size, - pci_hole64_size); -memory_region_init_alias (&f->pci_hole_64bit,OBJECT (d),"pci-hole64" , - f->pci_address_space, - i440fx->pci_info.w64.begin, - pci_hole64_size); -if (pci_hole64_size) { -memory_region_add_subregion (f->system_memory, - i440fx->pci_info.w64.begin, - &f->pci_hole_64bit); - } +/* setup pci memory mapping */ +pc_pci_as_mapping_init (OBJECT (f), f->system_memory, + f->pci_address_space); +memory_region_init_alias (&f->smram_region,OBJECT (d),"smram-region" , f->pci_address_space,0xa0000 ,0x20000 ); memory_region_add_subregion_overlap(f->system_memory, 0xa0000,
hw/pci-host/q35.c +5 -22
@@ -336,28 +336,11 @@ static int mch_init(PCIDevice *d) {int i;MCHPCIState *mch =MCH_PCI_DEVICE (d); -uint64_t pci_hole64_size; - -/* setup pci memory regions */ -memory_region_init_alias (&mch->pci_hole,OBJECT (mch),"pci-hole" , - mch->pci_address_space, - mch->below_4g_mem_size, -0x100000000ULL - mch->below_4g_mem_size); -memory_region_add_subregion (mch->system_memory, mch->below_4g_mem_size, - &mch->pci_hole); - - pci_hole64_size =pci_host_get_hole64_size (mch->pci_hole64_size); -pc_init_pci64_hole (&mch->pci_info,0x100000000ULL + mch->above_4g_mem_size, - pci_hole64_size); -memory_region_init_alias (&mch->pci_hole_64bit,OBJECT (mch),"pci-hole64" , - mch->pci_address_space, - mch->pci_info.w64.begin, - pci_hole64_size); -if (pci_hole64_size) { -memory_region_add_subregion (mch->system_memory, - mch->pci_info.w64.begin, - &mch->pci_hole_64bit); - } + +/* setup pci memory mapping */ +pc_pci_as_mapping_init (OBJECT (mch), mch->system_memory, + mch->pci_address_space); +/* smram */ cpu_smm_register (&mch_set_smm, mch); memory_region_init_alias(&mch->smram_region, OBJECT(mch), "smram-region",
include/hw/i386/pc.h +2 -12
@@ -108,17 +108,9 @@ PcGuestInfo *pc_guest_info_init(ram_addr_t below_4g_mem_size,#define PCI_HOST_PROP_PCI_HOLE64_SIZE "pci-hole64-size"#define DEFAULT_PCI_HOLE64_SIZE (~0x0ULL) -static inline uint64_t pci_host_get_hole64_size (uint64_t pci_hole64_size) -{ -if (pci_hole64_size ==DEFAULT_PCI_HOLE64_SIZE ) { -return 1ULL <<62 ; - }else { -return pci_hole64_size; - } -} -void pc_init_pci64_hole (PcPciInfo *pci_info,uint64_t pci_hole64_start, -uint64_t pci_hole64_size); +void pc_pci_as_mapping_init (Object *owner,MemoryRegion *system_memory, +MemoryRegion *pci_address_space); FWCfgState *pc_memory_init(MemoryRegion *system_memory, const char *kernel_filename,
@@ -167,8 +159,6 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state, int *piix_devfn, MemoryRegion *address_space_mem, MemoryRegion *address_space_io, ram_addr_t ram_size, - hwaddr pci_hole_start, - hwaddr pci_hole_size, ram_addr_t above_4g_mem_size, MemoryRegion *pci_memory, MemoryRegion *ram_memory);
include/hw/pci-host/q35.h +0 -2
@@ -53,8 +53,6 @@ typedef struct MCHPCIState {MemoryRegion *address_space_io;PAMMemoryRegion pam_regions[13 ];MemoryRegion smram_region; -MemoryRegion pci_hole; -MemoryRegion pci_hole_64bit;PcPciInfo pci_info;uint8_t smm_enabled;ram_addr_t below_4g_mem_size;
Signature · 1 line
1.7.1
PATCH v2 2/2 pc: add 'etc/pcimem64-minimum-address' fw_cfg interface to SeaBIOS
In reply to Igor Mammedov'etc/pcimem64-minimum-address' will allow QEMU to tell BIOS where PCI memory address space mapping could start in high memory.
Allowing BIOS to start mapping 64-bit PCI BARs at address where it wouldn't conflict with other mappings QEMU might place before it.
That permits QEMU to reserve extra address space before 64-bit PCI hole for memory hotplug.
Related SeaBIOS patch: http://patchwork.ozlabs.org/patch/283623/
--- hw/i386/pc.c | 14 +++++++++++++- hw/pci-host/piix.c | 3 ++- hw/pci-host/q35.c | 3 ++- include/hw/i386/pc.h | 3 ++- 4 files changed, 19 insertions(+), 4 deletions(-)
hw/i386/pc.c +13 -1
@@ -1055,11 +1055,23 @@ PcGuestInfo *pc_guest_info_init(ram_addr_t below_4g_mem_size,/* setup pci memory address space mapping into system address space */ void pc_pci_as_mapping_init (Object *owner,MemoryRegion *system_memory, -MemoryRegion *pci_address_space) +MemoryRegion *pci_address_space, +uint64_t pcimem64_min_addr) { +uint64_t *val; +FWCfgState *fw_cfg =fw_cfg_find (); +/* Set to lower priority than RAM */ memory_region_add_subregion_overlap (system_memory,0x0 , pci_address_space,-1 ); +g_assert (fw_cfg); +/* +* Align address at 1G, this makes sure it can be exactly covered +* with a PAT entry even when using huge pages. +*/ + val =g_malloc (sizeof (*val)); + *val =cpu_to_le64 (ROUND_UP (pcimem64_min_addr,0x1ULL <<30 )); +fw_cfg_add_file (fw_cfg,"etc/pcimem64-minimum-address" , val,sizeof (*val)); }void pc_acpi_init (const char *default_dsdt)
hw/pci-host/piix.c +2 -1
@@ -351,7 +351,8 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state,/* setup pci memory mapping */ pc_pci_as_mapping_init (OBJECT (f), f->system_memory, - f->pci_address_space); + f->pci_address_space, +0x100000000ULL + above_4g_mem_size);memory_region_init_alias (&f->smram_region,OBJECT (d),"smram-region" , f->pci_address_space,0xa0000 ,0x20000 );
hw/pci-host/q35.c +2 -1
@@ -339,7 +339,8 @@ static int mch_init(PCIDevice *d)/* setup pci memory mapping */ pc_pci_as_mapping_init (OBJECT (mch), mch->system_memory, - mch->pci_address_space); + mch->pci_address_space, +0x100000000ULL + mch->above_4g_mem_size);/* smram */ cpu_smm_register (&mch_set_smm, mch);
include/hw/i386/pc.h +2 -1
@@ -110,7 +110,8 @@ PcGuestInfo *pc_guest_info_init(ram_addr_t below_4g_mem_size,void pc_pci_as_mapping_init (Object *owner,MemoryRegion *system_memory, -MemoryRegion *pci_address_space); +MemoryRegion *pci_address_space, +uint64_t pcimem64_min_addr); FWCfgState *pc_memory_init(MemoryRegion *system_memory, const char *kernel_filename,
Signature · 1 line
1.7.1
Quoted 7 lines · Igor Mammedov
* simplify PCI address space mapping into system address space, replacing code duplication in piix/q53 PCs with a helper function
* add
fw_cfg'etc/pcimem64-minimum-address' to allow QEMU reserve additional address space before 64-bit PCI hole. Which will be need for reserving memory hotplug region in highmem. SeaBIOS counterpart: http://patchwork.ozlabs.org/patch/283623/
I'd like to see if we can figure out the migration issue with memory layout. Because if we do, and get rid of the separate 64 bit region as a concept, exposing the start of this non-existent region in FW CFG will make very little sense IMHO.
Quoted 24 lines
v2:
* use negative priority to map PCI address space under RAM memory regions which allows simplify code by removingpci_hole&pci_hole64memory region aliasesSeries depends on:
"memory: Change MemoryRegion priorities from unsigned to signed:Git tree for testing:
https://github.com/imammedo/qemu/commits/pcimem64-minimum-address-v2Igor Mammedov (1):
pc: add 'etc/pcimem64-minimum-address'fw_cfginterface to SeaBIOSMichael S. Tsirkin (1):
pc: map PCI address space as catchall region for not mapped addresseshw/i386/pc.c | 28 ++++++++++++++++------------ hw/i386/pc_piix.c | 2 -- hw/pci-host/piix.c | 27 +++++---------------------- hw/pci-host/q35.c | 28 ++++++---------------------- include/hw/i386/pc.h | 15 +++------------ include/hw/pci-host/q35.h | 2 -- 6 files changed, 30 insertions(+), 72 deletions(-)
Quoted 11 lines · On Tue, 29 Oct 2013 17:10:47 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 7 lines · Igor Mammedov
* simplify PCI address space mapping into system address space, replacing code duplication in piix/q53 PCs with a helper function
* add
fw_cfg'etc/pcimem64-minimum-address' to allow QEMU reserve additional address space before 64-bit PCI hole. Which will be need for reserving memory hotplug region in highmem. SeaBIOS counterpart: http://patchwork.ozlabs.org/patch/283623/I'd like to see if we can figure out the migration issue with memory layout.
It seems that there isn't migration issue here.
Quoted 3 lines
Because if we do, and get rid of the separate 64 bit region as a concept, exposing the start of this non-existent region in FW CFG will make very little sense IMHO.
Well, BIOS have to know where it could start 64-bit BARs mappings and telling it explicitly where, looks like a good way to do it.
Quoted 25 lines
Quoted 25 lines
v2:
* use negative priority to map PCI address space under RAM memory regions which allows simplify code by removingpci_hole&pci_hole64memory region aliasesSeries depends on:
"memory: Change MemoryRegion priorities from unsigned to signed:Git tree for testing:
https://github.com/imammedo/qemu/commits/pcimem64-minimum-address-v2Igor Mammedov (1):
pc: add 'etc/pcimem64-minimum-address'fw_cfginterface to SeaBIOSMichael S. Tsirkin (1):
pc: map PCI address space as catchall region for not mapped addresseshw/i386/pc.c | 28 ++++++++++++++++------------ hw/i386/pc_piix.c | 2 -- hw/pci-host/piix.c | 27 +++++---------------------- hw/pci-host/q35.c | 28 ++++++---------------------- include/hw/i386/pc.h | 15 +++------------ include/hw/pci-host/q35.h | 2 -- 6 files changed, 30 insertions(+), 72 deletions(-)
Quoted 7 lines · Igor Mammedov
Quoted 4 lines · On Tue, 29 Oct 2013 17:10:47 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 3 lines · Igor Mammedov
* simplify PCI address space mapping into system address space, replacing code duplication in piix/q53 PCs with a helper function
Quoted 8 lines
Quoted 7 lines
Quoted 4 lines
* add
fw_cfg'etc/pcimem64-minimum-address' to allow QEMU reserve additional address space before 64-bit PCI hole. Which will be need for reserving memory hotplug region in highmem. SeaBIOS counterpart: http://patchwork.ozlabs.org/patch/283623/I'd like to see if we can figure out the migration issue with memory layout.
It seems that there isn't migration issue here.
Hmm earlier you thought there was - it's ok now?
Quoted 6 lines
Quoted 3 lines
Because if we do, and get rid of the separate 64 bit region as a concept, exposing the start of this non-existent region in FW CFG will make very little sense IMHO.
Well, BIOS have to know where it could start 64-bit BARs mappings and
telling it explicitly where, looks like a good way to do it.
As far as I can tell, BIOS can start any mappings anywhere it wants to as long as they don't overlap anything else.
What is has to know is what hardware is there.
Quoted 25 lines
Quoted 25 lines
Quoted 25 lines
v2:
* use negative priority to map PCI address space under RAM memory regions which allows simplify code by removingpci_hole&pci_hole64memory region aliasesSeries depends on:
"memory: Change MemoryRegion priorities from unsigned to signed:Git tree for testing:
https://github.com/imammedo/qemu/commits/pcimem64-minimum-address-v2Igor Mammedov (1):
pc: add 'etc/pcimem64-minimum-address'fw_cfginterface to SeaBIOSMichael S. Tsirkin (1):
pc: map PCI address space as catchall region for not mapped addresseshw/i386/pc.c | 28 ++++++++++++++++------------ hw/i386/pc_piix.c | 2 -- hw/pci-host/piix.c | 27 +++++---------------------- hw/pci-host/q35.c | 28 ++++++---------------------- include/hw/i386/pc.h | 15 +++------------ include/hw/pci-host/q35.h | 2 -- 6 files changed, 30 insertions(+), 72 deletions(-)
Hi,
Quoted 7 lines
Quoted 3 lines
Well, BIOS have to know where it could start 64-bit BARs mappings and
telling it explicitly where, looks like a good way to do it.As far as I can tell, BIOS can start any mappings anywhere it wants to as long as they don't overlap anything else.
What is has to know is what hardware is there.
Use case is memory hotplug. Once we generate the acpi tables in qemu seabios doesn't need to know anything about hotpluggable memory slots. Still it better should not map 64bit pci bars into that address space. So it IMHO makes sense to give seabios a hit where it should map the 64bit pci bars.
cheers,
Gerd
Quoted 22 lines · On Tue, 29 Oct 2013 20:52:42 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 7 lines · Igor Mammedov
Quoted 4 lines · On Tue, 29 Oct 2013 17:10:47 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 3 lines · Igor Mammedov
* simplify PCI address space mapping into system address space, replacing code duplication in piix/q53 PCs with a helper function
Quoted 8 lines
Quoted 7 lines
Quoted 4 lines
* add
fw_cfg'etc/pcimem64-minimum-address' to allow QEMU reserve additional address space before 64-bit PCI hole. Which will be need for reserving memory hotplug region in highmem. SeaBIOS counterpart: http://patchwork.ozlabs.org/patch/283623/I'd like to see if we can figure out the migration issue with memory layout.
It seems that there isn't migration issue here.
Hmm earlier you thought there was - it's ok now?
Quoted 3 lines
Quoted 3 lines
Because if we do, and get rid of the separate 64 bit region as a concept, exposing the start of this non-existent region in FW CFG will make very little sense IMHO.
Quoted 7 lines
Quoted 3 lines
Well, BIOS have to know where it could start 64-bit BARs mappings and
telling it explicitly where, looks like a good way to do it.As far as I can tell, BIOS can start any mappings anywhere it wants to as long as they don't overlap anything else.
What is has to know is what hardware is there.
lets suppose we are describing HW to Seabios (about which it doesn't really needs to know when we move ACPI tables into QEMU)
1. we inform Seabios where memory hotplug region ends (describing new hw) and name parameter "etc/mem-hoplug-region-end" and make appropriate change to Seabois so it would know about memory hotlug region.
2. than in several releases we decide to add another device that would reserve address space after memory hotplug region. That would require another modification of Seabios to teach it about new hardware so it could place 64-bit PCI mappings after it.
So describing a new hardware each time will only increase amount of PV interfaces overtime, breaking old BIOSes and forcing us to maintain compatibility layers for old versions in Seabios and QEMU.
Opposite approach in this series tries to minimize all above mentioned problems by providing Seabios with an explicit information that it needs to make correct 64-bit PCI BARs mappings and it won't introduce compatibility issues with Seabios if we reserve extra space behind memory hotlpug region in future.
Quoted 28 lines
Quoted 27 lines
Quoted 27 lines
Quoted 27 lines
v2:
* use negative priority to map PCI address space under RAM memory regions which allows simplify code by removingpci_hole&pci_hole64memory region aliasesSeries depends on:
"memory: Change MemoryRegion priorities from unsigned to signed:Git tree for testing:
https://github.com/imammedo/qemu/commits/pcimem64-minimum-address-v2Igor Mammedov (1):
pc: add 'etc/pcimem64-minimum-address'fw_cfginterface to SeaBIOSMichael S. Tsirkin (1):
pc: map PCI address space as catchall region for not mapped addresseshw/i386/pc.c | 28 ++++++++++++++++------------ hw/i386/pc_piix.c | 2 -- hw/pci-host/piix.c | 27 +++++---------------------- hw/pci-host/q35.c | 28 ++++++---------------------- include/hw/i386/pc.h | 15 +++------------ include/hw/pci-host/q35.h | 2 -- 6 files changed, 30 insertions(+), 72 deletions(-)
Quoted 32 lines · Igor Mammedov
Quoted 22 lines · On Tue, 29 Oct 2013 20:52:42 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 7 lines · Igor Mammedov
Quoted 4 lines · On Tue, 29 Oct 2013 17:10:47 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 3 lines · Igor Mammedov
* simplify PCI address space mapping into system address space, replacing code duplication in piix/q53 PCs with a helper function
Quoted 8 lines
Quoted 7 lines
Quoted 4 lines
* add
fw_cfg'etc/pcimem64-minimum-address' to allow QEMU reserve additional address space before 64-bit PCI hole. Which will be need for reserving memory hotplug region in highmem. SeaBIOS counterpart: http://patchwork.ozlabs.org/patch/283623/I'd like to see if we can figure out the migration issue with memory layout.
It seems that there isn't migration issue here.
Hmm earlier you thought there was - it's ok now?
Quoted 3 lines
Quoted 3 lines
Because if we do, and get rid of the separate 64 bit region as a concept, exposing the start of this non-existent region in FW CFG will make very little sense IMHO.
Quoted 7 lines
Quoted 3 lines
Well, BIOS have to know where it could start 64-bit BARs mappings and
telling it explicitly where, looks like a good way to do it.As far as I can tell, BIOS can start any mappings anywhere it wants to as long as they don't overlap anything else.
What is has to know is what hardware is there.
Quoted 12 lines
lets suppose we are describing HW to Seabios (about which it doesn't really needs to know when we move ACPI tables into QEMU)
1. we inform Seabios where memory hotplug region ends (describing new hw) and name parameter "etc/mem-hoplug-region-end" and make appropriate change to Seabois so it would know about memory hotlug region.
2. than in several releases we decide to add another device that would reserve address space after memory hotplug region. That would require another modification of Seabios to teach it about new hardware so it could place 64-bit PCI mappings after it.
So describing a new hardware each time will only increase amount of PV interfaces overtime, breaking old BIOSes and forcing us to maintain compatibility layers for old versions in Seabios and QEMU.
I don't think we can predict the future.
It seems just as likely that BIOS will need to drive the
new hardware so it will not be enough to have "start pci here". In particular, non-contiguous hotpluggable memory does not seem all that far-fetched, and we might want to put PCI devices in the holes that are left free.
Maybe, if you want it make it kind of generic you can just say "reserved-memory-end" with the same value.
Quoted 6 lines
Opposite approach in this series tries to minimize all above mentioned problems by providing Seabios with an explicit information that it needs to make correct 64-bit PCI BARs mappings and it won't introduce compatibility issues with Seabios if we reserve extra space behind memory hotlpug region in future.
Point is that the value you expose has nothing to do with PCI. Yes BIOS might use it for that but interfaces need to tell what they are, not what they are used for.
Quoted 28 lines
Quoted 28 lines
Quoted 27 lines
Quoted 27 lines
Quoted 27 lines
v2:
* use negative priority to map PCI address space under RAM memory regions which allows simplify code by removingpci_hole&pci_hole64memory region aliasesSeries depends on:
"memory: Change MemoryRegion priorities from unsigned to signed:Git tree for testing:
https://github.com/imammedo/qemu/commits/pcimem64-minimum-address-v2Igor Mammedov (1):
pc: add 'etc/pcimem64-minimum-address'fw_cfginterface to SeaBIOSMichael S. Tsirkin (1):
pc: map PCI address space as catchall region for not mapped addresseshw/i386/pc.c | 28 ++++++++++++++++------------ hw/i386/pc_piix.c | 2 -- hw/pci-host/piix.c | 27 +++++---------------------- hw/pci-host/q35.c | 28 ++++++---------------------- include/hw/i386/pc.h | 15 +++------------ include/hw/pci-host/q35.h | 2 -- 6 files changed, 30 insertions(+), 72 deletions(-)
Hi,
Quoted 9 lines
I don't think we can predict the future.
It seems just as likely that BIOS will need to drive the
new hardware so it will not be enough to have "start pci here". In particular, non-contiguous hotpluggable memory does not seem all that far-fetched, and we might want to put PCI devices in the holes that are left free.Maybe, if you want it make it kind of generic you can just say "reserved-memory-end" with the same value.
What is the plan for handling e820 with memory hotplug btw?
You've mentioned e820 entries for hotpluggable dimms present at boot are needed for some guests.
I'm about to add a new e820 fw_cfg interface (see
http://comments.gmane.org/gmane.comp.emulators.qemu/238860).
So maybe design that with memory hotplug in mind? Such as adding a new qemu-specific type QEMU_RAM_HOTPLUG? Which seabios could use to reserve the memory (but not add it to the e820 table for the guest)?
cheers,
Gerd
Quoted 9 lines · On Wed, 30 Oct 2013 15:33:35 +0100 Gerd Hoffmann <[email protected]> wrote:
Hi,
Quoted 7 lines
I don't think we can predict the future.
It seems just as likely that BIOS will need to drive the
new hardware so it will not be enough to have "start pci here". In particular, non-contiguous hotpluggable memory does not seem all that far-fetched, and we might want to put PCI devices in the holes that are left free.
Quoted 4 lines
Quoted 2 lines
Maybe, if you want it make it kind of generic you can just say "reserved-memory-end" with the same value.
What is the plan for handling e820 with memory hotplug btw?
There wasn't immediate plans for using e820 for hotplugged memory yet.
Quoted 2 lines
You've mentioned e820 entries for hotpluggable dimms present at boot are needed for some guests.
Windows guest are fine and do not require e820 entries for present hotplugged memory, automatically onlining memory devices described in ACPI. Linux guests see memory devices described in ACPI but do not online them automatically.
I'm not sure how valid would be putting present hotplugged memory into e820 which is supposed to be static.
It would be better if linux guests behave as windows, it could online memory regions in userspace (using udev) or fix kernel to do it at boot time after scanning ACPI devices.
Quoted 6 lines
I'm about to add a new e820
fw_cfginterface (see
http://comments.gmane.org/gmane.comp.emulators.qemu/238860).So maybe design that with memory hotplug in mind? Such as adding a new qemu-specific type
QEMU_RAM_HOTPLUG? Which seabios could use to reserve the memory (but not add it to the e820 table for the guest)?
It will do job too. But extending semantics of standard table would be confusing. Yes, Seabios will filter it out but it doesn't make table less confusing.
I'd prefer having a dedicated interface for it as a more clean solution.
Quoted 4 lines
cheers,
Gerd
Hi,
Quoted 6 lines
Quoted 3 lines
So maybe design that with memory hotplug in mind? Such as adding a new qemu-specific type
QEMU_RAM_HOTPLUG? Which seabios could use to reserve the memory (but not add it to the e820 table for the guest)?It will do job too. But extending semantics of standard table would be confusing. Yes, Seabios will filter it out but it doesn't make table less confusing.
Was just thinking that it might be easier that way if we need e820 entries for hotplug memory address space _anyway_.
Quoted 1 line
I'd prefer having a dedicated interface for it as a more clean solution.
Agree.
cheers,
Gerd
Quoted 11 lines · On Mon, 04 Nov 2013 13:48:03 +0100 Gerd Hoffmann <[email protected]> wrote:
Hi,
Quoted 6 lines
Quoted 3 lines
So maybe design that with memory hotplug in mind? Such as adding a new qemu-specific type
QEMU_RAM_HOTPLUG? Which seabios could use to reserve the memory (but not add it to the e820 table for the guest)?It will do job too. But extending semantics of standard table would be confusing. Yes, Seabios will filter it out but it doesn't make table less confusing.
Was just thinking that it might be easier that way if we need e820 entries for hotplug memory address space _anyway_.
I don't think that we need e820 entries for hotplug memory reserved space as e820 should. In case present at boot hotpluggable DIMMs would be needed in E820, we can add them as usual E820_RAM entries.
But regardless of what we do here it might be good keep option of adding non standard entries in future, by filtering out unknown types in SeaBIOS.
Quoted 4 lines
Quoted 2 lines
I'd prefer having a dedicated interface for it as a more clean solution.
Agree.
So back to naming question, would you agree to renaming fw_cfg to the last Michael's suggestion "reserved-memory-end"?
Quoted 5 lines
cheers,
Gerd
Quoted 22 lines · Igor Mammedov
Quoted 11 lines · On Mon, 04 Nov 2013 13:48:03 +0100 Gerd Hoffmann <[email protected]> wrote:
Hi,
Quoted 6 lines
Quoted 3 lines
So maybe design that with memory hotplug in mind? Such as adding a new qemu-specific type
QEMU_RAM_HOTPLUG? Which seabios could use to reserve the memory (but not add it to the e820 table for the guest)?It will do job too. But extending semantics of standard table would be confusing. Yes, Seabios will filter it out but it doesn't make table less confusing.
Was just thinking that it might be easier that way if we need e820 entries for hotplug memory address space _anyway_.
I don't think that we need e820 entries for hotplug memory reserved space as e820 should. In case present at boot hotpluggable DIMMs would be needed in E820, we can add them as usual
E820_RAMentries.But regardless of what we do here it might be good keep option of adding non standard entries in future, by filtering out unknown types in SeaBIOS.
Quoted 1 line
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
I'd prefer having a dedicated interface for it as a more clean solution.
Agree.
So back to naming question, would you agree to renaming
fw_cfgto the last Michael's suggestion "reserved-memory-end"?
Fine with me.
cheers,
Gerd
Quoted 13 lines · On Tue, 29 Oct 2013 20:52:42 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 12 lines · Igor Mammedov
Quoted 9 lines · On Tue, 29 Oct 2013 17:10:47 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 7 lines · Igor Mammedov
* simplify PCI address space mapping into system address space, replacing code duplication in piix/q53 PCs with a helper function
* add
fw_cfg'etc/pcimem64-minimum-address' to allow QEMU reserve additional address space before 64-bit PCI hole. Which will be need for reserving memory hotplug region in highmem. SeaBIOS counterpart: http://patchwork.ozlabs.org/patch/283623/
Quoted 5 lines
Quoted 3 lines
Quoted 2 lines
I'd like to see if we can figure out the migration issue with memory layout.
It seems that there isn't migration issue here.
Hmm earlier you thought there was - it's ok now?
You've probably missed my reply
http://www.mail-archive.com/[email protected]/msg199586.html
it was error it test environment.