https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63926
Bug ID: 63926
Summary: [5 Regression] ICE in estimate_edge_growth, at
ipa-inline.h:300
Product: gcc
Version: 5.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: middle-end
Assignee: unassigned at gcc dot gnu.org
Reporter: Joost.VandeVondele at mat dot ethz.ch
I see the following ICE with recent trunk (appeared in the last week or so):
gfortran -c -fprofile-use -O3 -march=native -funroll-loops -ffast-math
-ffree-form -D__GFORTRAN -D__FFTSG -D__FFTW3 -D__LIBINT -D__HAS_smm_dnn
-D__COMPILE_ARCH="\"gfortran-test28\"" -D__COMPILE_DATE="\"Tue Nov 18 03:15:49
CET 2014\"" -D__COMPILE_HOST="\"nanosim-s01.ethz.ch\""
-D__COMPILE_REVISION="\"12221\"" -D__SHORT_FILE__="\"xc_xpbe_hole_t_c_lr.F\""
/data/vjoost/gnu/cp2k/cp2k/makefiles/../src/xc_xpbe_hole_t_c_lr.F
/data/vjoost/gnu/cp2k/cp2k/makefiles/../src/xc_xpbe_hole_t_c_lr.F:108:0:
END SUBROUTINE xpbe_hole_t_c_lr_lda_info
^
internal compiler error: in estimate_edge_growth, at ipa-inline.h:300
0x9a813c estimate_edge_growth
../../gcc/gcc/ipa-inline.h:299
0x9a813c estimate_edge_growth
../../gcc/gcc/ipa-inline-analysis.c:3854
0x9a813c do_estimate_growth_1
../../gcc/gcc/ipa-inline-analysis.c:3867
0x76792e cgraph_node::call_for_symbol_thunks_and_aliases(bool (*)(cgraph_node*,
void*), void*, bool)
../../gcc/gcc/cgraph.c:2206
0x9a2779 do_estimate_growth(cgraph_node*)
../../gcc/gcc/ipa-inline-analysis.c:3881
0x11f2f80 estimate_growth
../../gcc/gcc/ipa-inline.h:276
0x11f2f80 inline_small_functions
../../gcc/gcc/ipa-inline.c:1609
0x11f2f80 ipa_inline
../../gcc/gcc/ipa-inline.c:2198
0x11f2f80 execute
../../gcc/gcc/ipa-inline.c:2568
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.
If I copy source, objects, mod and gcda files to different directory, I run
into an error: Error: source locations for function
‘__xc_xpbe_hole_t_c_lr_MOD_xpbe_hole_t_c_lr_lda_info’ have changed, the profile
data may be out of date [-Werror=coverage-mismatch]. Is there an option to
allow for this (-Wno-error=coverage-mismatch results in no ICE)?
