Re: [PATCH] mm: Outline copy_overflow()

[Christophe Leroy]

Le 07/02/2022 à 21:40, Andrew Morton a écrit :
> On Mon,  7 Feb 2022 11:55:18 +0100 Christophe Leroy <christophe.leroy@xxxxxxxxxx> wrote:
> 
>> While building a small config with CONFIG_CC_OPTIMISE_FOR_SIZE,
>> I ended up with more than 50 times the following function in vmlinux:
>>
>> 	c00243bc <copy_overflow>:
>> 	c00243bc:	94 21 ff f0 	stwu    r1,-16(r1)
>> 	c00243c0:	7c 85 23 78 	mr      r5,r4
>> 	c00243c4:	7c 64 1b 78 	mr      r4,r3
>> 	c00243c8:	3c 60 c0 62 	lis     r3,-16286
>> 	c00243cc:	7c 08 02 a6 	mflr    r0
>> 	c00243d0:	38 63 5e e5 	addi    r3,r3,24293
>> 	c00243d4:	90 01 00 14 	stw     r0,20(r1)
>> 	c00243d8:	4b ff 82 45 	bl      c001c61c <__warn_printk>
>> 	c00243dc:	0f e0 00 00 	twui    r0,0
>> 	c00243e0:	80 01 00 14 	lwz     r0,20(r1)
>> 	c00243e4:	38 21 00 10 	addi    r1,r1,16
>> 	c00243e8:	7c 08 03 a6 	mtlr    r0
>> 	c00243ec:	4e 80 00 20 	blr
>>
>> That function being a non conditional warning on an error path,
>> it is not worth inlining.
>>
>> Outline it.
> 
> "uninline" is the conventional term for this.
> 
>> This reduces the size of vmlinux by almost 4kbytes.
> 
> Did you consider uninlining check_copy_size() instead?
> 

If you uninline check_copy_size() you loose the benefit of constant 
folding and build time overflow detection for constant sizes, and I'm 
not sure __builtin_object_size() works at all if uninlined.

Christophe