Skip to contentSkip to conversation

e820: pass high memory too.

RFC PATCH In review Gerd Hoffmann 2 participants 5 messages Oct 10, 2013 – Oct 11, 2013

lore
Share & tools
lore
#0 lore raw

RFC PATCH 1/1 e820: pass high memory too.

We have a fw_cfg entry to pass e820 entries from qemu to the firmware. Today it's used to pass reservations only. This patch makes qemu pass entries for RAM too.

This allows to pass RAM sizes larger than 1TB to the firmware and it will also allow to pass non-contignous memory ramges should we decide to implement that some day, say for our virtual numa nodes.

Obviously this needs some extra care to not break existing firware.

SeaBIOS loads the entries and happily adds them without looking at the type. Which is problematic for memory below 4g as this will overwrite reservations added for bios memory etc. For memory above 4g it works just fine, seabios will merge the entry derived from cmos with the one loaded from fw_cfg.

OVMF doesn't look at the fw_cfg e820 table.
coreboot doesn't look at the fw_cfg e820 table.

Cc Andrea Arcangeli <[email protected]>
---
 hw/i386/pc.c | 8 ++++++++
 1 file changed, 8 insertions(+)
hw/i386/pc.c +8 -0
index 0c313fe..ec5508b 100644
@@ -1134,12 +1134,20 @@ FWCfgState *pc_memory_init(MemoryRegion *system_memory,
     memory_region_init_alias(ram_below_4g, NULL, "ram-below-4g", ram,
                              0, below_4g_mem_size);
     memory_region_add_subregion(system_memory, 0, ram_below_4g);
+    if (0) {
+        /*
+         * Ideally we should do that too, but that would ruin the e820
+         * reservations added by seabios before initializing fw_cfg.
+         */
+        e820_add_entry(0, below_4g_mem_size, E820_RAM);
+    }
     if (above_4g_mem_size > 0) {
         ram_above_4g = g_malloc(sizeof(*ram_above_4g));
         memory_region_init_alias(ram_above_4g, NULL, "ram-above-4g", ram,
                                  below_4g_mem_size, above_4g_mem_size);
         memory_region_add_subregion(system_memory, 0x100000000ULL,
                                     ram_above_4g);
+        e820_add_entry(0x100000000ULL, above_4g_mem_size, E820_RAM);
     }
 
 
Signature · 1 line
1.8.3.1
#1 lore raw
Quoted 8 lines · On Thu, 10 Oct 2013 14:54:29 +0200 Gerd Hoffmann <[email protected]> wrote:

We have a fw_cfg entry to pass e820 entries from qemu to the firmware. Today it's used to pass reservations only. This patch makes qemu pass entries for RAM too.

This allows to pass RAM sizes larger than 1TB to the firmware and it will also allow to pass non-contignous memory ramges should we decide to implement that some day, say for our virtual numa nodes.

Quoted 7 lines

Obviously this needs some extra care to not break existing firware.

SeaBIOS loads the entries and happily adds them without looking at the type. Which is problematic for memory below 4g as this will overwrite reservations added for bios memory etc. For memory above 4g it works just fine, seabios will merge the entry derived from cmos with the one loaded from fw_cfg.

It will make amount of available memory in e820 table more than described in smbios and could break MS's SMBIOS HCT test. Perhaps related smbios info also should be picked up from QEMU.

Quoted 35 lines

OVMF doesn't look at the fw_cfg e820 table.
coreboot doesn't look at the fw_cfg e820 table.

Cc Andrea Arcangeli <[email protected]>
---
 hw/i386/pc.c | 8 ++++++++
 1 file changed, 8 insertions(+)
hw/i386/pc.c +8 -0
index 0c313fe..ec5508b 100644
@@ -1134,12 +1134,20 @@ FWCfgState *pc_memory_init(MemoryRegion *system_memory,
     memory_region_init_alias(ram_below_4g, NULL, "ram-below-4g", ram,
                              0, below_4g_mem_size);
     memory_region_add_subregion(system_memory, 0, ram_below_4g);
+    if (0) {
+        /*
+         * Ideally we should do that too, but that would ruin the e820
+         * reservations added by seabios before initializing fw_cfg.
+         */
+        e820_add_entry(0, below_4g_mem_size, E820_RAM);
+    }
     if (above_4g_mem_size > 0) {
         ram_above_4g = g_malloc(sizeof(*ram_above_4g));
         memory_region_init_alias(ram_above_4g, NULL, "ram-above-4g", ram,
                                  below_4g_mem_size, above_4g_mem_size);
         memory_region_add_subregion(system_memory, 0x100000000ULL,
                                     ram_above_4g);
+        e820_add_entry(0x100000000ULL, above_4g_mem_size, E820_RAM);
     }
 
 
#2 lore raw
Quoted 13 lines · Igor Mammedov
Quoted 10 lines · On Thu, 10 Oct 2013 14:54:29 +0200 Gerd Hoffmann <[email protected]> wrote:

We have a fw_cfg entry to pass e820 entries from qemu to the firmware. Today it's used to pass reservations only. This patch makes qemu pass entries for RAM too.

This allows to pass RAM sizes larger than 1TB to the firmware and it will also allow to pass non-contignous memory ramges should we decide to implement that some day, say for our virtual numa nodes.

Obviously this needs some extra care to not break existing firware.

Quoted 7 lines
Quoted 5 lines

SeaBIOS loads the entries and happily adds them without looking at the type. Which is problematic for memory below 4g as this will overwrite reservations added for bios memory etc. For memory above 4g it works just fine, seabios will merge the entry derived from cmos with the one loaded from fw_cfg.

It will make amount of available memory in e820 table more than described in smbios and could break MS's SMBIOS HCT test.

Happens only in case the installed amount of memory is larger than 1TB (which is the maximum the cmos can describe). Such a setup doesn't work correctly without the patch anyway, so it isn't a regression IMHO. And with an additional patch on the seabios side we can fix the smbios info even for the >1TB case.

Quoted 2 lines

Perhaps related smbios info
also should be picked up from QEMU.

Worth investigating, but lets concentrate on the acpi table merge first.

cheers,
Gerd

#3 lore raw
Quoted 19 lines · On Fri, 11 Oct 2013 14:27:12 +0200 Gerd Hoffmann <[email protected]> wrote:
Quoted 18 lines · Igor Mammedov
Quoted 15 lines · On Thu, 10 Oct 2013 14:54:29 +0200 Gerd Hoffmann <[email protected]> wrote:

We have a fw_cfg entry to pass e820 entries from qemu to the firmware. Today it's used to pass reservations only. This patch makes qemu pass entries for RAM too.

This allows to pass RAM sizes larger than 1TB to the firmware and it will also allow to pass non-contignous memory ramges should we decide to implement that some day, say for our virtual numa nodes.

Obviously this needs some extra care to not break existing firware.

SeaBIOS loads the entries and happily adds them without looking at the type. Which is problematic for memory below 4g as this will overwrite reservations added for bios memory etc. For memory above 4g it works just fine, seabios will merge the entry derived from cmos with the one loaded from fw_cfg.

Quoted 8 lines
Quoted 2 lines

It will make amount of available memory in e820 table more than described in smbios and could break MS's SMBIOS HCT test.

Happens only in case the installed amount of memory is larger than 1TB (which is the maximum the cmos can describe). Such a setup doesn't work correctly without the patch anyway, so it isn't a regression IMHO. And with an additional patch on the seabios side we can fix the smbios info even for the >1TB case.

True, and it looks better then adding extra CMOS ports.

BTW: what about OVMF and core boot, how would they know that there is more then 1Tb RAM?

Quoted 10 lines
Quoted 3 lines

Perhaps related smbios info
also should be picked up from QEMU.

Worth investigating, but lets concentrate on the acpi table merge first.

cheers,
Gerd

#4 lore raw

Hi,

Quoted 2 lines

BTW: what about OVMF and core boot, how would they know that there is more then 1Tb RAM?

They should look at fw_cfg too. Both have support for fw_cfg already, but don't parse the e820 table today. Shouldn't be that hard to add though.

cheers,
Gerd