Re: [PATCH 07/10] cmake (Windows): complain when encountering an unknown compiler
- From
Johannes Schindelin <johannes.schindelin@gmx.de>
- Date
- Sep 26, 2020, 20:33 UTC
- Message-ID
- <nycvar.QRO.7.76.6.2009262232560.50@tvgsbejvaqbjf.bet>
- In-Reply-To
- <CAKiG+9XqJ92rjVczBG9ByVK_KRfJeToO0pFb3RhFpom2ChFE-g@mail.gmail.com>
Hi Sibi,
On Fri, 25 Sep 2020, Sibi Siddharthan wrote:
Show 18 quoted lines
> On Fri, Sep 25, 2020 at 7:58 PM Johannes Schindelin via GitGitGadget > <gitgitgadget@gmail.com> wrote: > > > > From: Johannes Schindelin <johannes.schindelin@gmx.de> > > > > We have some custom handling regarding the link options, which are > > specific to each compiler. > > > > Therefore: let's not just continue without setting the link options when > > configuring for a currently unhandled compiler, but error out. > > > > Note: for some reason, the `invalidcontinue.obj` trick seems not to work > > in this developer's hands when building Git in Debug mode: the assertion > > dialog that is supposed to be suppressed still pops up, complaining > > about the `entry` variable being uninitialized when it is used in > > `config.c`'s line 1961. > > Does this happen when running Git under the Visual Studio debugger?
Whoops. I meant to remove this paragraph after addressing that `run-time check failure #3` problem.
Will remove it before sending the next iteration.
Thanks, Dscho
Show 11 quoted lines
> If so this might help > _CrtSetReportMode(_CRT_ASSERT,0); > > After digging a bit more I found out that _CRTDBG_MODE_DEBUG is > environment sensitive. > It depends on how the debugger handles DbgPrintEx(). > > Thank You, > Sibi Siddharthan > >