Ping again.
Thanks,
Feng
From: Feng Xue OS
Sent: Wednesday, September 1, 2021 9:46 AM
To: Martin Liška; Jan Hubicka; gcc@gcc.gnu.org
Cc: JiangNing OS
Subject: PING: [RFC] Whole Program Devirtualization
Honza,
How do you think about proposal in this RF
Honza,
How do you think about proposal in this RFC? Thanks a lot.
Best Regards,
Feng
From: Martin Liška
Sent: Friday, August 20, 2021 9:45 PM
To: Feng Xue OS; gcc@gcc.gnu.org
Cc: JiangNing OS; Jan Hubicka
Subject: Re: [RFC] Whole Program Devirtualizatio
We are not going to create a new devirtualization framework from
scratch, just hope it to be an enhancement on current speculative
devirtualization. The process does not need parse native code in
library, but only resort to existing lightweight symbol resolution
by LTO-prelinker. And C++ virtual di
1. Background
C++ Devirtualization is meant to determine all possible class type
identities of a virtual dynamic call, and try to transform it to direct
static call, or certain kind of statements with lower runtime cost.
Current implementation in GCC tends to adopt a somewhat conservative
scheme,
Sorry, sent to wrong mail list.
From: Feng Xue OS
Sent: Friday, May 14, 2021 4:30 PM
To: gcc-patc...@gcc.gnu.org
Subject: Question about non-POD class type
For an instance of a non-POD class, can I always assume that any
operation on it should be type-safe
>> To simplify explanation of this memory gathering optimization, pseudo
>> code on original nested loops is given as:
>>
>> outer-loop ( ) { /* data in inner loop are also invariant in outer loop.
>> */
>> ...
>> inner-loop (iter, iter_count) { /* inner loop to apply MGO */
>>
>>
>> -Original Message-
>> From: Feng Xue OS
>> Sent: Thursday, January 14, 2021 12:28 PM
>> To: gcc@gcc.gnu.org
>> Cc: JiangNing OS ; Hao Liu OS
>>
>> Subject: [RFC] A memory gathering optimization for loop
>>
>> 1. Background
>>
>> In a loop, it is optimal if only one memory stream is act
1. Background
In a loop, it is optimal if only one memory stream is activated, that
is, all memory operations sequentially access one data region. But that
is always not the case, such as traversing link list and manipulating
discrete arrays. In this scenario, the loop would contain multiple
scatt
1. Background
In a loop, it is optimal if only one memory stream is activated, that
is, all memory operations sequentially access one data region. But that
is always not the case, such as traversing link list and manipulating
discrete arrays. In this scenario, the loop would contain multiple
scatt
>>
>> >
>> >>
>> >> >> There is a match-folding issue derived from pr94234. A piece of
>> >> >> code like:
>> >> >>
>> >> >> int foo (int n)
>> >> >> {
>> >> >> int t1 = 8 * n;
>> >> >> int t2 = 8 * (n - 1);
>> >> >>
>> >> >> return t1 - t2;
>> >> >> }
>> >> >>
>> >> >> It
>
>>
>> >> There is a match-folding issue derived from pr94234. A piece of code
>> >> like:
>> >>
>> >> int foo (int n)
>> >> {
>> >> int t1 = 8 * n;
>> >> int t2 = 8 * (n - 1);
>> >>
>> >> return t1 - t2;
>> >> }
>> >>
>> >> It can be perfectly caught by the rule "(A * C)
>> There is a match-folding issue derived from pr94234. A piece of code like:
>>
>> int foo (int n)
>> {
>> int t1 = 8 * n;
>> int t2 = 8 * (n - 1);
>>
>> return t1 - t2;
>> }
>>
>> It can be perfectly caught by the rule "(A * C) +- (B * C) -> (A +- B) *
>> C", and
>>
Hi,
There is a match-folding issue derived from pr94234. A piece of code like:
int foo (int n)
{
int t1 = 8 * n;
int t2 = 8 * (n - 1);
return t1 - t2;
}
It can be perfectly caught by the rule "(A * C) +- (B * C) -> (A +- B) * C",
and
be folded to constant "8". But thi
13 matches
Mail list logo
