linux_dsm_epyc7002/arch/arm/mm/copypage-feroceon.c

113 lines
3.1 KiB
C
Raw Normal View History

/*
* linux/arch/arm/mm/copypage-feroceon.S
*
* Copyright (C) 2008 Marvell Semiconductors
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*
* This handles copy_user_highpage and clear_user_page on Feroceon
* more optimally than the generic implementations.
*/
#include <linux/init.h>
#include <linux/highmem.h>
static void __naked
feroceon_copy_user_page(void *kto, const void *kfrom)
{
asm("\
stmfd sp!, {r4-r9, lr} \n\
mov ip, %2 \n\
1: mov lr, r1 \n\
ldmia r1!, {r2 - r9} \n\
pld [lr, #32] \n\
pld [lr, #64] \n\
pld [lr, #96] \n\
pld [lr, #128] \n\
pld [lr, #160] \n\
pld [lr, #192] \n\
pld [lr, #224] \n\
stmia r0, {r2 - r9} \n\
ldmia r1!, {r2 - r9} \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
stmia r0, {r2 - r9} \n\
ldmia r1!, {r2 - r9} \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
stmia r0, {r2 - r9} \n\
ldmia r1!, {r2 - r9} \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
stmia r0, {r2 - r9} \n\
ldmia r1!, {r2 - r9} \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
stmia r0, {r2 - r9} \n\
ldmia r1!, {r2 - r9} \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
stmia r0, {r2 - r9} \n\
ldmia r1!, {r2 - r9} \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
stmia r0, {r2 - r9} \n\
ldmia r1!, {r2 - r9} \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
stmia r0, {r2 - r9} \n\
subs ip, ip, #(32 * 8) \n\
mcr p15, 0, r0, c7, c14, 1 @ clean and invalidate D line\n\
add r0, r0, #32 \n\
bne 1b \n\
mcr p15, 0, ip, c7, c10, 4 @ drain WB\n\
ldmfd sp!, {r4-r9, pc}"
:
: "r" (kto), "r" (kfrom), "I" (PAGE_SIZE));
}
void feroceon_copy_user_highpage(struct page *to, struct page *from,
unsigned long vaddr, struct vm_area_struct *vma)
{
void *kto, *kfrom;
kto = kmap_atomic(to);
kfrom = kmap_atomic(from);
ARM: Flush user mapping on VIVT processors when copying a page Steven Walter <stevenrwalter@gmail.com> writes: > I've been tracking down an instance of userspace data corruption, > and I believe I have found a window during fork where data can be > lost. The corruption is occurring on an ARMv5 system with VIVT > caches. Here's the scenario in question. Thread A is forking, > Thread B is running in userspace: > > Thread A: flush_cache_mm() (dup_mmap) > Thread B: writes to a page in the above mm > Thread A: pte_wrprotect() the above page (copy_one_pte) > Thread B: writes to the same page again > > During thread B's second write, he'll take a fault and enter the > do_wp_page() case. We'll end up calling copy_page(), which notably > uses the kernel virtual addresses for the old and new pages. This > means that the new page does not necessarily have the data from the > first write. Now there are two conflicting copies of the same > cache-line in dcache. If the userspace cache-line flushes before > the kernel cache-line, we lose the changes made during the first > write. do_wp_page does call flush_dcache_page on the newly-copied > page, but there's still a window where the CPU could flush the > userspace cache-line before then. Resolve this by flushing the user mapping before copying the page on processors with a writeback VIVT cache. Note: this does have a performance impact, and so needs further consideration before being merged - can we optimize out some of the cache flushes if, eg, we know that the page isn't yet mapped? Thread: <e06498070903061426o5875ad13hc6328aa0d3f08ed7@mail.gmail.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2009-10-05 21:34:22 +07:00
flush_cache_page(vma, vaddr, page_to_pfn(from));
feroceon_copy_user_page(kto, kfrom);
kunmap_atomic(kfrom);
kunmap_atomic(kto);
}
void feroceon_clear_user_highpage(struct page *page, unsigned long vaddr)
{
void *ptr, *kaddr = kmap_atomic(page);
asm volatile ("\
mov r1, %2 \n\
mov r2, #0 \n\
mov r3, #0 \n\
mov r4, #0 \n\
mov r5, #0 \n\
mov r6, #0 \n\
mov r7, #0 \n\
mov ip, #0 \n\
mov lr, #0 \n\
1: stmia %0, {r2-r7, ip, lr} \n\
subs r1, r1, #1 \n\
mcr p15, 0, %0, c7, c14, 1 @ clean and invalidate D line\n\
add %0, %0, #32 \n\
bne 1b \n\
mcr p15, 0, r1, c7, c10, 4 @ drain WB"
: "=r" (ptr)
: "0" (kaddr), "I" (PAGE_SIZE / 32)
: "r1", "r2", "r3", "r4", "r5", "r6", "r7", "ip", "lr");
kunmap_atomic(kaddr);
}
struct cpu_user_fns feroceon_user_fns __initdata = {
.cpu_clear_user_highpage = feroceon_clear_user_highpage,
.cpu_copy_user_highpage = feroceon_copy_user_highpage,
};