[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: feature/igc: pdump segment never freed
From: |
Daniel Colascione |
Subject: |
Re: feature/igc: pdump segment never freed |
Date: |
Thu, 22 May 2025 11:12:30 -0400 |
User-agent: |
K-9 Mail for Android |
On May 22, 2025 3:47:44 AM EDT, Eli Zaretskii <eliz@gnu.org> wrote:
>> Date: Thu, 22 May 2025 06:51:47 +0000
>> Cc: Gerd Möllmann <gerd.moellmann@gmail.com>,
>> Emacs Devel <emacs-devel@gnu.org>
>> From: Pip Cet via "Emacs development discussions." <emacs-devel@gnu.org>
>>
>> I'd appreciate more ideas, but right now I think for many users the best
>> thing we can do is to load the dump one object at a time, fix up the
>> relocations, then drop the temporary data. It works, but it does appear
>> to mean some tens of milliseconds for every Emacs lauMöllmann
>Uh-oh, I envision complaints from people who start Emacs frequently
>for some reason.
>
It's also just giving up. Other systems with fancy GCs, (e.g. Android ART) have
image support, so it's not like it's unheard of. Are we going to vendor MPS? If
so, instead of accepting performance losses or contorting memory management, we
could just teach MPS how to deal with the mapped dump area as a heap segment.