https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69725
Martin Sebor <msebor at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|RESOLVED |UNCONFIRMED
Resolution|INVALID |---
--- Comment #6 from Martin Sebor <msebor at gcc dot gnu.org> ---
Thank you for the suggestion.
I tried --with-build-config=bootstrap-lto but then forgot to remove -flto from
BOOT_CFLAGS and got the following error (also mentioned in my comment #1 on bug
65313 where I used --with-build-config=bootstrap-lto without setting
BOOT_CFLAGS):
get_d.c: In function '__gmpn_get_d':
get_d.c:490:1: error: the control flow of function '__gmpn_get_d' does not
match its profile data (counter 'arcs') [-Werror=coverage-mismatch]
}
^
get_d.c:490:1: error: the control flow of function '__gmpn_get_d' does not
match its profile data (counter 'time_profiler') [-Werror=coverage-mismatch]
cc1: some warnings being treated as errors
I'll see if removing -flto from BOOT_CFLAGS lets the x86_64 build succeed.
However, the manual says this much about --with-build-config=bootstrap-lto:
Enables Link-Time Optimization for host tools during bootstrapping.
'BUILD_CONFIG=bootstrap-lto' is equivalent to adding -flto to 'BOOT_CFLAGS'.
If there is a difference then either the manual should be corrected or the
build machinery fixed. I don't know enough to decide which but I see others
using the BOOT_CFLAGS approach (see for example bug 61448) so I am reopening
this bug to make sure it gets fixed one way or the other.
