This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Improving reload inheritance code generation and predictability
- From: Richard Henderson <rth at redhat dot com>
- To: Bernd Schmidt <bernds at codesourcery dot com>
- Cc: Jeff Law <law at redhat dot com>, gcc-patches <gcc-patches at gcc dot gnu dot org>
- Date: Thu, 18 Nov 2010 11:43:37 -0800
- Subject: Re: Improving reload inheritance code generation and predictability
- References: <4CE53C3D.6020001@redhat.com> <4CE57FC8.1070705@codesourcery.com>
On 11/18/2010 11:34 AM, Bernd Schmidt wrote:
> On 11/18/2010 03:46 PM, Jeff Law wrote:
>> Basically we start searching forward in the stream from the current insn
>> needing a reload noting uses of spill regs as we go.
>
> Isn't this quadratic in the number of insns?
It depends on how soon we find all of the spills.
If we fail to find them all, for every insn, then yes quadratic
on the size of the block. If there's 1 spill per insn and we
always find it at the next insn, then it's linear.
Perhaps Jeff has some data on how much scanning we do in practice?
r~