Messages in this thread Patch in this message | | | Date | Sun, 24 Feb 2008 18:46:05 +0100 | From | Alexander van Heukelum <> | Subject | [PATCH] Fix alignment of early reservation for EBDA |
| |
Hi Andi,
My eyes fell on the following table in the boot messages:
early res: 0 [0-fff] BIOS data page early res: 1 [6000-7fff] SMP_TRAMPOLINE early res: 2 [200000-374557] TEXT DATA BSS early res: 3 [9fc00-a0bff] EBDA early res: 4 [8000-afff] PGTABLE
The memory reserved for the EBDA overflows into the area normally reserved for the VGA adaptor. It seems that you wanted to force the allocation to cover whole pages, like:
early res: 3 [9f000-9ffff] EBDA
This is what this patch implements.
Is it really necessary to force the allocation to a page boundary?
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>
arch/x86/kernel/head64.c | 12 +++++++----- 1 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index ad24408..2c52543 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -53,7 +53,7 @@ static void __init copy_bootdata(char *real_mode_data) static __init void reserve_ebda(void) { - unsigned ebda_addr, ebda_size; + unsigned int ebda_addr, ebda_size, ebda_end; /* * there is a real-mode segmented pointer pointing to the @@ -70,12 +70,14 @@ static __init void reserve_ebda(void) /* Round EBDA up to pages */ if (ebda_size == 0) ebda_size = 1; + if (ebda_size > 64) + ebda_size = 64; ebda_size <<= 10; - ebda_size = round_up(ebda_size + (ebda_addr & ~PAGE_MASK), PAGE_SIZE); - if (ebda_size > 64*1024) - ebda_size = 64*1024; + + ebda_end = ALIGN(ebda_addr + ebda_size, PAGE_SIZE); + ebda_addr = ALIGN(ebda_addr - PAGE_SIZE, PAGE_SIZE); - reserve_early(ebda_addr, ebda_addr + ebda_size, "EBDA"); + reserve_early(ebda_addr, ebda_end, "EBDA"); } void __init x86_64_start_kernel(char * real_mode_data)
| |