[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20230123065746.GB30529@lst.de>
Date: Mon, 23 Jan 2023 07:57:46 +0100
From: Christoph Hellwig <hch@....de>
To: Song Liu <song@...nel.org>
Cc: Christoph Hellwig <hch@....de>,
Thomas Gleixner <tglx@...utronix.de>,
Luis Chamberlain <mcgrof@...nel.org>,
Christophe Leroy <christophe.leroy@...roup.eu>,
songliubraving@...com, Peter Zijlstra <peterz@...radead.org>,
linux-modules@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH/RFC] module: replace module_layout with module_memory
On Fri, Jan 20, 2023 at 09:42:26AM -0800, Song Liu wrote:
> So here are the two versions, both with secondary addr_[min|max] for
> CONFIG_ARCH_WANTS_MODULES_DATA_IN_VMALLOC.
>
> v2. Just use mod_mem array:
> https://git.kernel.org/pub/scm/linux/kernel/git/song/md.git/commit/?h=remotes/song-md/new_module_alloc_build_test_v2
>
> v3. mod_mem array and the defines:
> #define mod_core_text mod_mem[MOD_MEM_TYPE_TEXT]
> #define mod_core_data mod_mem[MOD_MEM_TYPE_DATA]
> etc.
> https://git.kernel.org/pub/scm/linux/kernel/git/song/md.git/commit/?h=remotes/song-md/new_module_alloc_build_test_v3
I find v2 much preferably. Having magic layouts with different
access methods is horribly confusing. It might be a way changes
things if modifying every caller would be intrusive, but starting
out that way does not seem like a very good idea.
Powered by blists - more mailing lists