@@ -155,6 +155,8 @@ int build_e820_table(struct e820entry *e820,
{
unsigned int nr = 0, i, j;
uint32_t low_mem_end = hvm_info->low_mem_pgend << PAGE_SHIFT;
+ unsigned long firmware_mem_end =
+ RESERVED_MEMORY_DYNAMIC_START + (mem_mfns_allocated() << PAGE_SHIFT);
if ( !lowmem_reserved_base )
lowmem_reserved_base = 0xA0000;
@@ -199,8 +201,19 @@ int build_e820_table(struct e820entry *e820,
nr++;
/*
+ * Mark populated reserved memory that contains ACPI and other tables as
+ * ACPI NVS (non-reclaimable) space - that should help the guest to treat
+ * it correctly later (e.g. pass to the next kernel on kexec).
+ */
+
+ e820[nr].addr = RESERVED_MEMBASE;
+ e820[nr].size = firmware_mem_end - RESERVED_MEMBASE;
+ e820[nr].type = E820_NVS;
+ nr++;
+
+ /*
* Explicitly reserve space for special pages.
- * This space starts at RESERVED_MEMBASE an extends to cover various
+ * This space starts after ACPI region and extends to cover various
* fixed hardware mappings (e.g., LAPIC, IOAPIC, default SVGA framebuffer).
*
* If igd_opregion_pgbase we need to split the RESERVED region in two.
@@ -210,8 +223,8 @@ int build_e820_table(struct e820entry *e820,
{
uint32_t igd_opregion_base = igd_opregion_pgbase << PAGE_SHIFT;
- e820[nr].addr = RESERVED_MEMBASE;
- e820[nr].size = (uint32_t) igd_opregion_base - RESERVED_MEMBASE;
+ e820[nr].addr = firmware_mem_end;
+ e820[nr].size = igd_opregion_base - firmware_mem_end;
e820[nr].type = E820_RESERVED;
nr++;
@@ -227,7 +240,7 @@ int build_e820_table(struct e820entry *e820,
}
else
{
- e820[nr].addr = RESERVED_MEMBASE;
+ e820[nr].addr = firmware_mem_end;
e820[nr].size = (uint32_t)-e820[nr].addr;
e820[nr].type = E820_RESERVED;
nr++;
@@ -444,6 +444,12 @@ void mem_hole_populate_ram(xen_pfn_t mfn, uint32_t nr_mfns)
static uint32_t alloc_up = RESERVED_MEMORY_DYNAMIC_START - 1;
static uint32_t alloc_down = RESERVED_MEMORY_DYNAMIC_END;
+uint32_t mem_mfns_allocated(void)
+{
+ return (alloc_up >> PAGE_SHIFT) -
+ ((RESERVED_MEMORY_DYNAMIC_START - 1) >> PAGE_SHIFT);
+}
+
xen_pfn_t mem_hole_alloc(uint32_t nr_mfns)
{
alloc_down -= nr_mfns << PAGE_SHIFT;
@@ -200,6 +200,9 @@ void mem_hole_populate_ram(xen_pfn_t mfn, uint32_t nr_mfns);
/* Allocate a memory hole below 4GB. */
xen_pfn_t mem_hole_alloc(uint32_t nr_mfns);
+/* Return number of pages allocated */
+uint32_t mem_mfns_allocated(void);
+
/* Allocate memory in a reserved region below 4GB. */
void *mem_alloc(uint32_t size, uint32_t align);
#define virt_to_phys(v) ((unsigned long)(v))
Guest kernel does need to know in some cases where the tables are located to treat these regions properly. One example is kexec process where the first kernel needs to pass firmware region locations to the second kernel which is now a requirement after 02a3e3cdb7f12 ("x86/boot: Parse SRAT table and count immovable memory regions"). The memory that hvmloader allocates in the reserved region mostly contains these useful tables and could be safely indicated as ACPI without the need to designate a sub-region specially for that. Making it non-reclaimable (ACPI NVS) in contrast with regular ACPI (ACPI table) memory would avoid potential reuse of this memory by the guest taking into account this region may contain runtime structures like VM86 TSS, etc. If necessary, those can be moved away later and the region marked as reclaimable. Signed-off-by: Igor Druzhinin <igor.druzhinin@citrix.com> --- tools/firmware/hvmloader/e820.c | 21 +++++++++++++++++---- tools/firmware/hvmloader/util.c | 6 ++++++ tools/firmware/hvmloader/util.h | 3 +++ 3 files changed, 26 insertions(+), 4 deletions(-)