aboutsummaryrefslogtreecommitdiffstats
path: root/include/asm-generic/vmlinux.lds.h
diff options
context:
space:
mode:
authorTejun Heo <tj@kernel.org>2009-01-13 06:41:35 -0500
committerIngo Molnar <mingo@elte.hu>2009-01-16 08:19:46 -0500
commit1a51e3a0aed18767cf2762e95456ecfeb0bca5e6 (patch)
tree2d930218ef1072a59f7dac0f97bb03aa02796c8c /include/asm-generic/vmlinux.lds.h
parentc8f3329a0ddd751241e96b4100df7eda14b2cbc6 (diff)
x86: fold pda into percpu area on SMP
[ Based on original patch from Christoph Lameter and Mike Travis. ] Currently pdas and percpu areas are allocated separately. %gs points to local pda and percpu area can be reached using pda->data_offset. This patch folds pda into percpu area. Due to strange gcc requirement, pda needs to be at the beginning of the percpu area so that pda->stack_canary is at %gs:40. To achieve this, a new percpu output section macro - PERCPU_VADDR_PREALLOC() - is added and used to reserve pda sized chunk at the start of the percpu area. After this change, for boot cpu, %gs first points to pda in the data.init area and later during setup_per_cpu_areas() gets updated to point to the actual pda. This means that setup_per_cpu_areas() need to reload %gs for CPU0 while clearing pda area for other cpus as cpu0 already has modified it when control reaches setup_per_cpu_areas(). This patch also removes now unnecessary get_local_pda() and its call sites. A lot of this patch is taken from Mike Travis' "x86_64: Fold pda into per cpu area" patch. Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Diffstat (limited to 'include/asm-generic/vmlinux.lds.h')
-rw-r--r--include/asm-generic/vmlinux.lds.h25
1 files changed, 24 insertions, 1 deletions
diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index fc2f55f2dcd6..e53319cf29cb 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -441,9 +441,10 @@
441 . = __per_cpu_load + SIZEOF(.data.percpu); 441 . = __per_cpu_load + SIZEOF(.data.percpu);
442 442
443/** 443/**
444 * PERCPU_VADDR - define output section for percpu area 444 * PERCPU_VADDR_PREALLOC - define output section for percpu area with prealloc
445 * @vaddr: explicit base address (optional) 445 * @vaddr: explicit base address (optional)
446 * @phdr: destination PHDR (optional) 446 * @phdr: destination PHDR (optional)
447 * @prealloc: the size of prealloc area
447 * 448 *
448 * Macro which expands to output section for percpu area. If @vaddr 449 * Macro which expands to output section for percpu area. If @vaddr
449 * is not blank, it specifies explicit base address and all percpu 450 * is not blank, it specifies explicit base address and all percpu
@@ -455,11 +456,33 @@
455 * section in the linker script will go there too. @phdr should have 456 * section in the linker script will go there too. @phdr should have
456 * a leading colon. 457 * a leading colon.
457 * 458 *
459 * If @prealloc is non-zero, the specified number of bytes will be
460 * reserved at the start of percpu area. As the prealloc area is
461 * likely to break alignment, this macro puts areas in increasing
462 * alignment order.
463 *
458 * This macro defines three symbols, __per_cpu_load, __per_cpu_start 464 * This macro defines three symbols, __per_cpu_load, __per_cpu_start
459 * and __per_cpu_end. The first one is the vaddr of loaded percpu 465 * and __per_cpu_end. The first one is the vaddr of loaded percpu
460 * init data. __per_cpu_start equals @vaddr and __per_cpu_end is the 466 * init data. __per_cpu_start equals @vaddr and __per_cpu_end is the
461 * end offset. 467 * end offset.
462 */ 468 */
469#define PERCPU_VADDR_PREALLOC(vaddr, segment, prealloc) \
470 PERCPU_PROLOG(vaddr) \
471 . += prealloc; \
472 *(.data.percpu) \
473 *(.data.percpu.shared_aligned) \
474 *(.data.percpu.page_aligned) \
475 PERCPU_EPILOG(segment)
476
477/**
478 * PERCPU_VADDR - define output section for percpu area
479 * @vaddr: explicit base address (optional)
480 * @phdr: destination PHDR (optional)
481 *
482 * Macro which expands to output section for percpu area. Mostly
483 * identical to PERCPU_VADDR_PREALLOC(@vaddr, @phdr, 0) other than
484 * using slighly different layout.
485 */
463#define PERCPU_VADDR(vaddr, phdr) \ 486#define PERCPU_VADDR(vaddr, phdr) \
464 PERCPU_PROLOG(vaddr) \ 487 PERCPU_PROLOG(vaddr) \
465 *(.data.percpu.page_aligned) \ 488 *(.data.percpu.page_aligned) \