MTRR+SMBIOS Bochs patches
PATCH 1/1 MTRR+SMBIOS Bochs patches
Attached patch adds a series of patches against Bochs, from KVM tree:
- SMBIOS updates
- MTRR support
- unconditional mptable
Alex Williamson (4):
kvm: bios: extend MTRRs to above 4G
kvm: bios: cleanup/consolidate above 4G memory parsing
kvm: bios: update SMBIOS table to report memory above 4G
kvm: bios: switch MTRRs to cover only the PCI range and default to WB
Avi Kivity (3):
kvm: bios: generate mptable unconditionally
kvm: bios: add mtrr support
kvm: bios: smp mtrr support
Bill Rieske (2):
kvm: bios: resolve memory device roll over reporting issues with >32 guests
kvm: bios: fix smbios memory device length boundary condition
bios-pq/0002_smbios_report_4g.patch +32 -0
@@ -0,0 +1,32 @@ +kvm: bios: update SMBIOS table to report memory above 4G + +Signed-off-by: Alex Williamson <[email protected]> +Signed-off-by: Avi Kivity <[email protected]> + +From: Alex Williamson <[email protected]> + +Index: bochs/bios/rombios32.c +=================================================================== +--- bochs.orig/bios/rombios32.c ++++ bochs/bios/rombios32.c +@@ -2081,7 +2081,8 @@ void smbios_init(void) + { + unsigned cpu_num, nr_structs = 0, max_struct_size = 0; + char *start, *p, *q; +- int memsize = ram_size / (1024 * 1024); ++ int memsize = (ram_end == ram_size) ? ram_size / (1024 * 1024) : ++ (ram_end - (1ull << 32) + ram_size) / (1024 * 1024); + + #ifdef BX_USE_EBDA_TABLES + ebda_cur_addr = align(ebda_cur_addr, 16); +@@ -2108,8 +2109,8 @@ void smbios_init(void) + add_struct(smbios_type_4_init(p, cpu_num)); + add_struct(smbios_type_16_init(p, memsize)); + add_struct(smbios_type_17_init(p, memsize)); +- add_struct(smbios_type_19_init(p, memsize)); +- add_struct(smbios_type_20_init(p, memsize)); ++ add_struct(smbios_type_19_init(p, ram_end / (1024 * 1024))); ++ add_struct(smbios_type_20_init(p, ram_end / (1024 * 1024))); + add_struct(smbios_type_32_init(p)); + add_struct(smbios_type_127_init(p)); +
bios-pq/0003_mptable_unconditional.patch +23 -0
@@ -0,0 +1,23 @@ +kvm: bios: generate mptable unconditionally + +VMware ESX requires an mptable even for uniprocessor guests. + +Signed-off-by: Avi Kivity <[email protected]> +From: Avi Kivity <[email protected]> + +Index: bochs/bios/rombios32.c +=================================================================== +--- bochs.orig/bios/rombios32.c ++++ bochs/bios/rombios32.c +@@ -970,11 +970,6 @@ static void mptable_init(void) + int ioapic_id, i, len; + int mp_config_table_size; + +-#ifdef BX_QEMU +- if (smp_cpus <= 1) +- return; +-#endif +- + #ifdef BX_USE_EBDA_TABLES + mp_config_table = (uint8_t *)(ram_size - ACPI_DATA_SIZE - MPTABLE_MAX_SIZE); + #else
bios-pq/0004_mtrr_up.patch +119 -0
119 lines not on this page · raw message
bios-pq/0005_mtrr_smp.patch +125 -0
125 lines not on this page · raw message
bios-pq/0006_mtrr_above_4g.patch +61 -0
61 lines not on this page · raw message
bios-pq/0007_mtrr_above_4g_clean.patch +57 -0
57 lines not on this page · raw message
bios-pq/0008_mtrr_default.patch +52 -0
52 lines not on this page · raw message
bios-pq/0009_smbios_16gb.patch +183 -0
183 lines not on this page · raw message
bios-pq/0010_smbios_16g_boundary.patch +22 -0
@@ -0,0 +1,22 @@ +kvm: bios: fix smbios memory device length boundary condition + +dev_memsize ends up 0 when it shouldn't be on 16G boundary conditions. + +Signed-off-by: Bill Rieske <[email protected]> +Signed-off-by: Avi Kivity <[email protected]> + +From: Bill Rieske <[email protected]> + +Index: bochs/bios/rombios32.c +=================================================================== +--- bochs.orig/bios/rombios32.c ++++ bochs/bios/rombios32.c +@@ -2218,7 +2218,7 @@ void smbios_init(void) + for ( i = 0; i < nr_mem_devs; i++ ) + { + uint32_t dev_memsize = ((i == (nr_mem_devs - 1)) +- ? (memsize & 0x3fff) : 0x4000); ++ ? (((memsize-1) & 0x3fff)+1) : 0x4000); + add_struct(smbios_type_17_init(p, dev_memsize, i)); + add_struct(smbios_type_19_init(p, dev_memsize, i)); + add_struct(smbios_type_20_init(p, dev_memsize, i));
bios-pq/series +9 -0
@@ -1 +1,10 @@ 0001_bx-qemu.patch +0002_smbios_report_4g.patch +0003_mptable_unconditional.patch +0004_mtrr_up.patch +0005_mtrr_smp.patch +0006_mtrr_above_4g.patch +0007_mtrr_above_4g_clean.patch +0008_mtrr_default.patch +0009_smbios_16gb.patch +0010_smbios_16g_boundary.patch
Quoted 11 lines · Marcelo Tosatti
Attached patch adds a series of patches against Bochs, from KVM tree:
- SMBIOS updates
- MTRR support
- unconditional mptableAlex Williamson (4):
kvm: bios: extend MTRRs to above 4G
kvm: bios: cleanup/consolidate above 4G memory parsing
kvm: bios: update SMBIOS table to report memory above 4G
kvm: bios: switch MTRRs to cover only the PCI range and default to WB
Quoted 10 lines
Avi Kivity (3):
kvm: bios: generate mptable unconditionally
kvm: bios: add mtrr support
kvm: bios: smp mtrr supportBill Rieske (2):
kvm: bios: resolve memory device roll over reporting issues with >32 guests
kvm: bios: fix smbios memory device length boundary condition
Wow, awesome!
Can you send each patch as a member of a patch series so they can be reviewed individually? Please CC bochs-devel. I'll take care of putting them in the right place in SVN.
Regards,
Anthony Liguori
Quoted 21 lines · Anthony Liguori
Quoted 20 lines · Marcelo Tosatti
Attached patch adds a series of patches against Bochs, from KVM tree:
- SMBIOS updates
- MTRR support
- unconditional mptableAlex Williamson (4):
kvm: bios: extend MTRRs to above 4G
kvm: bios: cleanup/consolidate above 4G memory parsing
kvm: bios: update SMBIOS table to report memory above 4G
kvm: bios: switch MTRRs to cover only the PCI range and default
to WBAvi Kivity (3):
kvm: bios: generate mptable unconditionally
kvm: bios: add mtrr support
kvm: bios: smp mtrr supportBill Rieske (2):
kvm: bios: resolve memory device roll over reporting issues withQuoted 1 line
32 guests
Quoted 8 lines
Quoted 3 lines
kvm: bios: fix smbios memory device length boundary condition
Wow, awesome!
Can you send each patch as a member of a patch series so they can be reviewed individually? Please CC bochs-devel. I'll take care of putting them in the right place in SVN.
All MTRR settings will be completely ignored by Qemu (writes will be discarded, reads will be zero) unless you also apply my Qemu MTRR support patch.
Subject: Re: [Qemu-devel] [PATCH] MTRR support on x86, part 1
Date: Thu, 11 Dec 2008 23:37:05 +0100
Should I resend that patch? It still applies to Qemu svn HEAD.
Regards,
Carl-Daniel
Signature · 1 line
http://www.hailfinger.org/