No clue about MSYS but I build off the GCC 5 git branch under MSYS2
without any issues pretty easily.
On Sun, Aug 9, 2015 at 6:02 AM, Javier Jaramago Fernandez
wrote:
> Hi there! Thanks you so much for taking the time of attending me. I have
> been trying for several hours to compile the SVN vers
Using -flto-partition=1to1 or -flto-partition=none generally works
around these issues. I typically use 1to1 when using lto for anything
since I can rarely even get gcc to bootstrap correctly with lto
without it.
On Thu, May 21, 2015 at 8:09 AM, lh_mouse wrote:
> FWIW, you can try moving virtual
The patch below on the gcc-patches ML allows bootstrapping to go
through for mingw i686 dwarf2-eh builds:
https://gcc.gnu.org/ml/gcc-patches/2015-05/msg01212.html
It could probably use someone else's eyes on it.
On Mon, May 11, 2015 at 9:28 AM, Matt Breedlove
wrote:
> I assumed as
posted under. Has anyone reported this? I haven't
seen anything in gcc's bugtraq that looks relevant.
On Mon, May 11, 2015 at 8:16 AM, David Macek wrote:
> On 11. 5. 2015 13:52, Matt Breedlove wrote:
>> Also, I'm experiencing a currently unresolved issue building gcc
&
I've posted an issue (and one bug) to GCC regarding native lto
bootstrap issues with 5.1/5.1.1 (perhaps trunk also) in reference to
stpcpy and sys_siglist which can be found here:
https://gcc.gnu.org/ml/gcc/2015-05/msg00084.html
Essentially, slim lto bootstraps fails due to GCC swapping out the
s
Yes, I am. One thing to note that I saw was a patch committed to
msys2's mingw64 builds (though I didn't need it to build mine).
Link:
https://github.com/Alexpux/MINGW-packages/blob/master/mingw-w64-gcc/310-gcc-make-xmmintrin-header-cplusplus-compatible.patch
On Fri, May 1, 2015 at 8:53 AM, niXm
error: in read_cgraph_and_symbols, at lto/lto.c:2947" error and seems
to impact 4.9 although non-fatally. I'll be following both lists so
if any questions need answering, feel free to ask on whichever list
seems most appropriate
