x86: implement true end_pfn_mapped for 32bit
Even on 32bit 2MB pages can map more memory than is in the true
max_low_pfn if end_pfn is not highmem and not aligned to 2MB.
Add a end_pfn_map similar to x86-64 that accounts for this
fact. This is important for code that really needs to know about
all mapping aliases.
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: andreas.herrmann3@amd.com
Cc: mingo@elte.hu
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c
index 00168e6..73dd060 100644
--- a/arch/x86/mm/init_32.c
+++ b/arch/x86/mm/init_32.c
@@ -51,6 +51,8 @@
unsigned int __VMALLOC_RESERVE = 128 << 20;
+unsigned long end_pfn_map;
+
DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
unsigned long highstart_pfn, highend_pfn;
@@ -194,6 +196,7 @@
set_pmd(pmd, pfn_pmd(pfn, prot));
pfn += PTRS_PER_PTE;
+ end_pfn_map = pfn;
continue;
}
pte = one_page_table_init(pmd);
@@ -208,6 +211,7 @@
set_pte(pte, pfn_pte(pfn, prot));
}
+ end_pfn_map = pfn;
}
}
}