On 2022/7/27 14:59, Xi Ruoyao wrote:
On Tue, 2022-07-26 at 21:10 +0800, WANG Xuerui wrote:
On 2022/7/26 21:02, Huacai Chen wrote:
Disable executable stack for LoongArch by default, as all modern
architectures do.
I don't know why this slipped in under everyone's eyes... Struggling
to
recall some of my mental activities during the initial review, I may
be
not too familiar with the code at that time (maybe still the case
now),
and didn't check what exactly "read_implies_exec" means in this
particular context. That could be just the reason for my part.
But better mention the discussion leading to the discovery of this
bug:
"The problematic behavior was initially discovered by Andreas Schwab
in
a binutils discussion [1], fix suggested by WANG Xuerui" or something
along the line.
[1]: https://sourceware.org/pipermail/binutils/2022-July/121992.html
I think we already have a "standard" format for this:
"Reported-by: Andreas Schwab <...>"
"Suggested-by: Wang Xuerui <...>"
"Url: https://sourceware.org/pipermail/binutils/2022-July/121992.html";
Hmm, I didn't find any occurrence of "Url:" but plenty of "Link:" in the
kernel's commit history, so I agree the tag trio should sound even better...
Tested on my A2101 board and nothing is broken so far.
Tested-by: Xi Ruoyao <xry111@xxxxxxxxxxx>
--
WANG "xen0n" Xuerui
Linux/LoongArch mailing list: https://lore.kernel.org/loongarch/
