On Fri, 16 Dec 2022 16:10:10 GMT, Magnus Ihse Bursie wrote:
>> Justin King has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Simplify logic for including __ubsan_default_options
>>
>> Signed-off-by: Justin King
>
> I much also
On Fri, 16 Dec 2022 15:56:44 GMT, Magnus Ihse Bursie wrote:
> However, I do think the included source files should be treated like the
> autoheaders, and reside in data rather than in `src`. The latter is intended
> for buildtools, even though they are a bit scattered at the moment (there is
On Mon, 12 Dec 2022 10:42:22 GMT, Justin King wrote:
>> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
>> the image due to lots of undefined behavior (it invokes the built JVM).
>> Follow up PRs will either replace the undefined behavior with well defined
>>
On Fri, 9 Dec 2022 20:38:26 GMT, Erik Joelsson wrote:
>> make/autoconf/jdk-options.m4 line 450:
>>
>>> 448:
>>> ###
>>> 449: #
>>> 450: # UndefinedBehaviorSanitizer
>>
>> I think this logic fits better in `flags.m4`,
On Mon, 12 Dec 2022 10:42:22 GMT, Justin King wrote:
>> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
>> the image due to lots of undefined behavior (it invokes the built JVM).
>> Follow up PRs will either replace the undefined behavior with well defined
>>
On Tue, 13 Dec 2022 16:55:09 GMT, Robbin Ehn wrote:
> > I guess the advantage to putting this in the build machinery (as opposed to
> > using `--with-extra-cflags=-fsanitize=undefined
> > --with-extra-ldflags=-fsanitize=undefined`) is that we can turn some of
> > these onn by default once
On Tue, 13 Dec 2022 13:45:04 GMT, Justin King wrote:
> Nope. Some targets end up passing C++ flags to the C compiler, causing a
> failure.
Ah right, we (mis)use CFLAGS (instead of CXXFLAGS) in some
SetupNativeCompilation calls when all source files are C++. In that case, your
suggested patch
On Tue, 13 Dec 2022 16:29:59 GMT, Justin King wrote:
> I guess the advantage to putting this in the build machinery (as opposed to
> using `--with-extra-cflags=-fsanitize=undefined
> --with-extra-ldflags=-fsanitize=undefined`) is that we can turn some of these
> onn by default once we've
On Mon, 12 Dec 2022 10:42:22 GMT, Justin King wrote:
>> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
>> the image due to lots of undefined behavior (it invokes the built JVM).
>> Follow up PRs will either replace the undefined behavior with well defined
>>
On Mon, 12 Dec 2022 10:42:22 GMT, Justin King wrote:
>> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
>> the image due to lots of undefined behavior (it invokes the built JVM).
>> Follow up PRs will either replace the undefined behavior with well defined
>>
On Mon, 12 Dec 2022 10:42:22 GMT, Justin King wrote:
>> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
>> the image due to lots of undefined behavior (it invokes the built JVM).
>> Follow up PRs will either replace the undefined behavior with well defined
>>
On Tue, 13 Dec 2022 08:17:09 GMT, Justin King wrote:
> I tried to use a single file, but the build logic attempts to compile as
> either C or C++ based on file extensions, and has logic based on it. So if I
> use `.cpp` and the target is all `.c` odd things happen. The same for the
> inverse.
On Mon, 12 Dec 2022 10:42:22 GMT, Justin King wrote:
>> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
>> the image due to lots of undefined behavior (it invokes the built JVM).
>> Follow up PRs will either replace the undefined behavior with well defined
>>
On Mon, 12 Dec 2022 10:42:22 GMT, Justin King wrote:
>> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
>> the image due to lots of undefined behavior (it invokes the built JVM).
>> Follow up PRs will either replace the undefined behavior with well defined
>>
> Allow building OpenJDK with UBSan. Currently the build fails when optimizing
> the image due to lots of undefined behavior (it invokes the built JVM).
> Follow up PRs will either replace the undefined behavior with well defined
> behavior or suppress errors which are intentional. The goal is
15 matches
Mail list logo
