Re: [PATCH v3 4/4] powerpc/inst: Optimise copy_inst_from_kernel_nofault()

From: Christophe Leroy
Date: Mon Nov 29 2021 - 02:24:36 EST




Le 27/11/2021 à 19:10, Christophe Leroy a écrit :
copy_inst_from_kernel_nofault() uses copy_from_kernel_nofault() to
copy one or two 32bits words. This means calling an out-of-line
function which itself calls back copy_from_kernel_nofault_allowed()
then performs a generic copy with loops.

Rewrite copy_inst_from_kernel_nofault() to do everything at a
single place and use __get_kernel_nofault() directly to perform
single accesses without loops.

Before the patch:

00000018 <copy_inst_from_kernel_nofault>:
18: 94 21 ff e0 stwu r1,-32(r1)
1c: 7c 08 02 a6 mflr r0
20: 38 a0 00 04 li r5,4
24: 93 e1 00 1c stw r31,28(r1)
28: 7c 7f 1b 78 mr r31,r3
2c: 38 61 00 08 addi r3,r1,8
30: 90 01 00 24 stw r0,36(r1)
34: 48 00 00 01 bl 34 <copy_inst_from_kernel_nofault+0x1c>
34: R_PPC_REL24 copy_from_kernel_nofault
38: 2c 03 00 00 cmpwi r3,0
3c: 40 82 00 0c bne 48 <copy_inst_from_kernel_nofault+0x30>
40: 81 21 00 08 lwz r9,8(r1)
44: 91 3f 00 00 stw r9,0(r31)
48: 80 01 00 24 lwz r0,36(r1)
4c: 83 e1 00 1c lwz r31,28(r1)
50: 38 21 00 20 addi r1,r1,32
54: 7c 08 03 a6 mtlr r0
58: 4e 80 00 20 blr

After the patch:

00000018 <copy_inst_from_kernel_nofault>:
18: 3d 20 b0 00 lis r9,-20480
1c: 7c 04 48 40 cmplw r4,r9
20: 7c 69 1b 78 mr r9,r3
24: 41 80 00 2c blt 50 <copy_inst_from_kernel_nofault+0x38>
28: 81 42 04 d0 lwz r10,1232(r2)
2c: 39 4a 00 01 addi r10,r10,1
30: 91 42 04 d0 stw r10,1232(r2)
34: 80 e4 00 00 lwz r7,0(r4)
38: 81 42 04 d0 lwz r10,1232(r2)
3c: 38 60 00 00 li r3,0
40: 39 4a ff ff addi r10,r10,-1
44: 91 42 04 d0 stw r10,1232(r2)
48: 90 e9 00 00 stw r7,0(r9)
4c: 4e 80 00 20 blr

50: 38 60 ff de li r3,-34
54: 4e 80 00 20 blr
58: 81 22 04 d0 lwz r9,1232(r2)
5c: 38 60 ff f2 li r3,-14
60: 39 29 ff ff addi r9,r9,-1
64: 91 22 04 d0 stw r9,1232(r2)
68: 4e 80 00 20 blr

Signed-off-by: Christophe Leroy <christophe.leroy@xxxxxxxxxx>
---
v3: New
---
arch/powerpc/mm/maccess.c | 18 ++++++++++++------
1 file changed, 12 insertions(+), 6 deletions(-)

diff --git a/arch/powerpc/mm/maccess.c b/arch/powerpc/mm/maccess.c
index 5abae96b2b46..90309806f5eb 100644
--- a/arch/powerpc/mm/maccess.c
+++ b/arch/powerpc/mm/maccess.c
@@ -15,16 +15,22 @@ bool copy_from_kernel_nofault_allowed(const void *unsafe_src, size_t size)
int copy_inst_from_kernel_nofault(ppc_inst_t *inst, u32 *src)
{
unsigned int val, suffix;
- int err;
- err = copy_from_kernel_nofault(&val, src, sizeof(val));
- if (err)
- return err;
+ if (unlikely(!is_kernel_addr((unsigned long)src)))
+ return -ERANGE;
+
+ pagefault_disable();

Allthough generic version of copy_from_kernel_nofault() does it, disabling pagefault is pointless here because we are accessing kernel addresses only, so a page fault will always fail via bad_kernel_fault(), it will never reach the faulthandler_disabled() test.

+ __get_kernel_nofault(&val, src, u32, Efault);
if (IS_ENABLED(CONFIG_PPC64) && get_op(val) == OP_PREFIX) {
- err = copy_from_kernel_nofault(&suffix, src + 1, sizeof(suffix));
+ __get_kernel_nofault(&suffix, src + 1, u32, Efault);
+ pagefault_enable();
*inst = ppc_inst_prefix(val, suffix);
} else {
+ pagefault_enable();
*inst = ppc_inst(val);
}
- return err;
+ return 0;
+Efault:
+ pagefault_enable();
+ return -EFAULT;
}