spack/lib/spack/env/cc

544 lines
16 KiB
Plaintext
Raw Normal View History

#!/bin/bash
##############################################################################
# Copyright (c) 2013-2018, Lawrence Livermore National Security, LLC.
# Produced at the Lawrence Livermore National Laboratory.
#
# This file is part of Spack.
# Created by Todd Gamblin, tgamblin@llnl.gov, All rights reserved.
# LLNL-CODE-647188
#
# For details, see https://github.com/spack/spack
# Please also see the NOTICE and LICENSE files for our notice and the LGPL.
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU Lesser General Public License (as
# published by the Free Software Foundation) version 2.1, February 1999.
#
# This program is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the IMPLIED WARRANTY OF
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the terms and
# conditions of the GNU Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public
# License along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
##############################################################################
#
# Spack compiler wrapper script.
#
# Compiler commands go through this compiler wrapper in Spack builds.
# The compiler wrapper is a thin layer around the standard compilers.
# It enables several key pieces of functionality:
#
# 1. It allows Spack to swap compilers into and out of builds easily.
# 2. It adds several options to the compile line so that spack
# packages can find their dependencies at build time and run time:
# -I arguments for dependency /include directories.
# -L arguments for dependency /lib directories.
# -Wl,-rpath arguments for dependency /lib directories.
#
# This is an array of environment variables that need to be set before
2016-04-05 04:00:09 +08:00
# the script runs. They are set by routines in spack.build_environment
# as part of spack.package.Package.do_install().
parameters=(
SPACK_PREFIX
SPACK_ENV_PATH
SPACK_DEBUG_LOG_DIR
SPACK_DEBUG_LOG_ID
SPACK_COMPILER_SPEC
SPACK_CC_RPATH_ARG
SPACK_CXX_RPATH_ARG
SPACK_F77_RPATH_ARG
SPACK_FC_RPATH_ARG
SPACK_SHORT_SPEC
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
SPACK_SYSTEM_DIRS
)
# The compiler input variables are checked for sanity later:
# SPACK_CC, SPACK_CXX, SPACK_F77, SPACK_FC
# The default compiler flags are passed from these variables:
# SPACK_CFLAGS, SPACK_CXXFLAGS, SPACK_FCFLAGS, SPACK_FFLAGS,
# SPACK_LDFLAGS, SPACK_LDLIBS
# Debug env var is optional; set to "TRUE" for debug logging:
# SPACK_DEBUG
# Test command is used to unit test the compiler script.
# SPACK_TEST_COMMAND
# Dependencies can be empty for pkgs with no deps:
# SPACK_DEPENDENCIES
# die()
# Prints a message and exits with error 1.
function die {
echo "$@"
exit 1
}
# read system directories into an array (delimited by :)
IFS=':' read -ra SPACK_SYSTEM_DIRS <<< "${SPACK_SYSTEM_DIRS}"
# test whether a path is a system directory
function system_dir {
path="$1"
for sd in "${SPACK_SYSTEM_DIRS[@]}"; do
if [ "${path}" == "${sd}" ] || [ "${path}" == "${sd}/" ]; then
# success if path starts with a system prefix
return 0
fi
done
return 1 # fail if path starts no system prefix
}
for param in "${parameters[@]}"; do
2016-03-18 09:53:13 +08:00
if [[ -z ${!param} ]]; then
2016-04-05 19:42:23 +08:00
die "Spack compiler must be run from Spack! Input '$param' is missing."
fi
done
# Figure out the type of compiler, the language, and the mode so that
# the compiler script knows what to do.
#
# Possible languages are C, C++, Fortran 77, and Fortran 90.
# 'command' is set based on the input command to $SPACK_[CC|CXX|F77|F90]
#
# 'mode' is set to one of:
2016-04-05 04:00:09 +08:00
# vcheck version check
2016-03-18 09:53:13 +08:00
# cpp preprocess
# cc compile
2016-03-18 09:53:13 +08:00
# as assemble
# ld link
# ccld compile & link
command=$(basename "$0")
comp="CC"
case "$command" in
2016-04-05 02:33:48 +08:00
cpp)
mode=cpp
;;
cc|c89|c99|gcc|clang|icc|pgcc|xlc|xlc_r)
command="$SPACK_CC"
language="C"
comp="CC"
lang_flags=C
;;
c++|CC|g++|clang++|icpc|pgc++|xlc++|xlc++_r)
command="$SPACK_CXX"
language="C++"
comp="CXX"
lang_flags=CXX
;;
ftn|f90|fc|f95|gfortran|flang|ifort|pgfortran|xlf90|xlf90_r|nagfor)
command="$SPACK_FC"
language="Fortran 90"
comp="FC"
lang_flags=F
;;
f77|xlf|xlf_r|pgf77)
command="$SPACK_F77"
language="Fortran 77"
comp="F77"
lang_flags=F
;;
ld)
mode=ld
;;
*)
die "Unkown compiler: $command"
;;
esac
2016-04-05 04:00:09 +08:00
# If any of the arguments below are present, then the mode is vcheck.
# In vcheck mode, nothing is added in terms of extra search paths or
2016-04-05 02:33:48 +08:00
# libraries.
if [[ -z $mode ]] || [[ $mode == ld ]]; then
for arg in "$@"; do
case $arg in
-v|-V|--version|-dumpversion)
mode=vcheck
break
;;
esac
done
fi
# Finish setting up the mode.
2016-04-05 02:33:48 +08:00
if [[ -z $mode ]]; then
mode=ccld
for arg in "$@"; do
2016-04-05 03:38:21 +08:00
if [[ $arg == -E ]]; then
mode=cpp
break
2016-04-05 03:38:21 +08:00
elif [[ $arg == -S ]]; then
2016-03-18 09:53:13 +08:00
mode=as
break
2016-04-05 03:38:21 +08:00
elif [[ $arg == -c ]]; then
mode=cc
break
fi
done
fi
# Set up rpath variable according to language.
eval rpath=\$SPACK_${comp}_RPATH_ARG
# Dump the mode and exit if the command is dump-mode.
2016-04-05 03:38:21 +08:00
if [[ $SPACK_TEST_COMMAND == dump-mode ]]; then
echo "$mode"
exit
fi
# Check that at least one of the real commands was actually selected,
# otherwise we don't know what to execute.
2016-03-18 09:53:13 +08:00
if [[ -z $command ]]; then
die "ERROR: Compiler '$SPACK_COMPILER_SPEC' does not support compiling $language programs."
fi
#
# Set paths as defined in the 'environment' section of the compiler config
# names are stored in SPACK_ENV_TO_SET
# values are stored in SPACK_ENV_SET_<varname>
#
IFS=':' read -ra env_set_varnames <<< "$SPACK_ENV_TO_SET"
for varname in "${env_set_varnames[@]}"; do
spack_varname="SPACK_ENV_SET_$varname"
export "$varname"="${!spack_varname}"
unset "$spack_varname"
done
2016-05-19 07:50:50 +08:00
#
# Filter '.' and Spack environment directories out of PATH so that
# this script doesn't just call itself
#
IFS=':' read -ra env_path <<< "$PATH"
IFS=':' read -ra spack_env_dirs <<< "$SPACK_ENV_PATH"
spack_env_dirs+=("" ".")
export PATH=""
2016-05-19 07:50:50 +08:00
for dir in "${env_path[@]}"; do
addpath=true
for env_dir in "${spack_env_dirs[@]}"; do
if [[ "$dir" == "$env_dir" ]]; then
2016-05-19 07:50:50 +08:00
addpath=false
break
fi
done
if $addpath; then
export PATH="${PATH:+$PATH:}$dir"
2016-05-19 07:50:50 +08:00
fi
done
2016-04-05 02:33:48 +08:00
if [[ $mode == vcheck ]]; then
exec "${command}" "$@"
2016-03-18 09:53:13 +08:00
fi
2016-04-05 02:33:48 +08:00
# Darwin's linker has a -r argument that merges object files together.
# It doesn't work with -rpath.
# This variable controls whether they are added.
add_rpaths=true
if [[ ($mode == ld || $mode == ccld) && "$SPACK_SHORT_SPEC" =~ "darwin" ]];
then
2016-04-05 02:33:48 +08:00
for arg in "$@"; do
if [[ ($arg == -r && $mode == ld) ||
($arg == -r && $mode == ccld) ||
($arg == -Wl,-r && $mode == ccld) ]]; then
2016-04-05 02:33:48 +08:00
add_rpaths=false
break
2016-04-05 03:38:21 +08:00
fi
2016-04-05 02:33:48 +08:00
done
fi
# Save original command for debug logging
input_command="$*"
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
#
# Parse the command line arguments.
#
# We extract -L, -I, and -Wl,-rpath arguments from the command line and
# recombine them with Spack arguments later. We parse these out so that
# we can make sure that system paths come last, that package arguments
# come first, and that Spack arguments are injected properly.
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
#
# All other arguments, including -l arguments, are treated as
# 'other_args' and left in their original order. This ensures that
# --start-group, --end-group, and other order-sensitive flags continue to
# work as the caller expects.
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
#
# The libs variable is initialized here for completeness, and it is also
# used later to inject flags supplied via `ldlibs` on the command
# line. These come into the wrappers via SPACK_LDLIBS.
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
#
includes=()
libdirs=()
rpaths=()
system_includes=()
system_libdirs=()
system_rpaths=()
libs=()
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
other_args=()
while [ -n "$1" ]; do
# an RPATH to be added after the case statement.
rp=""
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
case "$1" in
-I*)
arg="${1#-I}"
if [ -z "$arg" ]; then shift; arg="$1"; fi
if system_dir "$arg"; then
system_includes+=("$arg")
else
includes+=("$arg")
fi
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
;;
-L*)
arg="${1#-L}"
if [ -z "$arg" ]; then shift; arg="$1"; fi
if system_dir "$arg"; then
system_libdirs+=("$arg")
else
libdirs+=("$arg")
fi
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
;;
-l*)
arg="${1#-l}"
if [ -z "$arg" ]; then shift; arg="$1"; fi
other_args+=("-l$arg")
;;
-Wl,*)
arg="${1#-Wl,}"
if [ -z "$arg" ]; then shift; arg="$1"; fi
if [[ "$arg" = -rpath=* ]]; then
rp="${arg#-rpath=}"
elif [[ "$arg" = -rpath,* ]]; then
rp="${arg#-rpath,}"
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
elif [[ "$arg" = -rpath ]]; then
shift; arg="$1"
if [[ "$arg" != -Wl,* ]]; then
die "-Wl,-rpath was not followed by -Wl,*"
fi
rp="${arg#-Wl,}"
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
else
other_args+=("-Wl,$arg")
fi
;;
-Xlinker,*)
arg="${1#-Xlinker,}"
if [ -z "$arg" ]; then shift; arg="$1"; fi
if [[ "$arg" = -rpath=* ]]; then
rp="${arg#-rpath=}"
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
elif [[ "$arg" = -rpath ]]; then
shift; arg="$1"
if [[ "$arg" != -Xlinker,* ]]; then
die "-Xlinker,-rpath was not followed by -Xlinker,*"
fi
rp="${arg#-Xlinker,}"
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
else
other_args+=("-Xlinker,$arg")
fi
;;
-Xlinker)
if [[ "$2" == "-rpath" ]]; then
if [[ "$3" != "-Xlinker" ]]; then
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
die "-Xlinker,-rpath was not followed by -Xlinker,*"
fi
shift 3;
rp="$1"
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
else
other_args+=("$1")
fi
;;
*)
other_args+=("$1")
;;
esac
# test rpaths against system directories in one place.
if [ -n "$rp" ]; then
if system_dir "$rp"; then
system_rpaths+=("$rp")
else
rpaths+=("$rp")
fi
fi
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
shift
done
#
# Add flags from Spack's cppflags, cflags, cxxflags, fcflags, fflags, and
# ldflags. We stick to the order that gmake puts the flags in by default.
#
# See the gmake manual on implicit rules for details:
# https://www.gnu.org/software/make/manual/html_node/Implicit-Variables.html
#
flags=()
# Fortran flags come before CPPFLAGS
case "$mode" in
cc|ccld)
case $lang_flags in
F)
flags=("${flags[@]}" "${SPACK_FFLAGS[@]}") ;;
esac
;;
esac
# C preprocessor flags come before any C/CXX flags
case "$mode" in
cpp|as|cc|ccld)
flags=("${flags[@]}" "${SPACK_CPPFLAGS[@]}") ;;
esac
# Add C and C++ flags
case "$mode" in
cc|ccld)
case $lang_flags in
C)
flags=("${flags[@]}" "${SPACK_CFLAGS[@]}") ;;
CXX)
flags=("${flags[@]}" "${SPACK_CXXFLAGS[@]}") ;;
esac
;;
esac
# Linker flags
case "$mode" in
ld|ccld)
flags=("${flags[@]}" "${SPACK_LDFLAGS[@]}") ;;
esac
# Include the package's prefix/lib[64] dirs in rpath. We don't know until
# *after* installation which one's correct, so we include both lib and
# lib64, assuming that only one will be present.
case "$mode" in
ld|ccld)
$add_rpaths && rpaths+=("$SPACK_PREFIX/lib")
$add_rpaths && rpaths+=("$SPACK_PREFIX/lib64")
;;
esac
# Read spack dependencies from the environment. This is a list of prefixes.
IFS=':' read -ra deps <<< "$SPACK_DEPENDENCIES"
for dep in "${deps[@]}"; do
# Append include directories in any compilation mode
case "$mode" in
cpp|cc|as|ccld)
if [[ -d $dep/include ]]; then
includes=("${includes[@]}" "$dep/include")
fi
;;
esac
# Append lib/lib64 and RPATH directories, but only if we're linking
case "$mode" in
ld|ccld)
if [[ -d $dep/lib ]]; then
if [[ $SPACK_RPATH_DEPS == *$dep* ]]; then
$add_rpaths && rpaths=("${rpaths[@]}" "$dep/lib")
fi
if [[ $SPACK_LINK_DEPS == *$dep* ]]; then
libdirs=("${libdirs[@]}" "$dep/lib")
fi
fi
if [[ -d $dep/lib64 ]]; then
if [[ $SPACK_RPATH_DEPS == *$dep* ]]; then
$add_rpaths && rpaths+=("$dep/lib64")
fi
if [[ $SPACK_LINK_DEPS == *$dep* ]]; then
libdirs+=("$dep/lib64")
fi
fi
;;
esac
done
# add RPATHs if we're in in any linking mode
case "$mode" in
ld|ccld)
# Set extra RPATHs
IFS=':' read -ra extra_rpaths <<< "$SPACK_COMPILER_EXTRA_RPATHS"
for extra_rpath in "${extra_rpaths[@]}"; do
$add_rpaths && rpaths+=("$extra_rpath")
libdirs+=("$extra_rpath")
done
# Add SPACK_LDLIBS to args
for lib in "${SPACK_LDLIBS[@]}"; do
libs+=("${lib#-l}")
done
;;
esac
#
# Finally, reassemble the command line.
#
# Includes and system includes first
args=()
# flags assembled earlier
args+=("${flags[@]}")
# include directory search paths
for dir in "${includes[@]}"; do args+=("-I$dir"); done
for dir in "${system_includes[@]}"; do args+=("-I$dir"); done
# Library search paths
for dir in "${libdirs[@]}"; do args+=("-L$dir"); done
for dir in "${system_libdirs[@]}"; do args+=("-L$dir"); done
# RPATHs arguments
case "$mode" in
ccld)
for dir in "${rpaths[@]}"; do args+=("$rpath$dir"); done
for dir in "${system_rpaths[@]}"; do args+=("$rpath$dir"); done
;;
ld)
for dir in "${rpaths[@]}"; do args+=("-rpath" "$dir"); done
for dir in "${system_rpaths[@]}"; do args+=("-rpath" "$dir"); done
;;
esac
# Other arguments from the input command
args+=("${other_args[@]}")
# Inject SPACK_LDLIBS, if supplied
for lib in "${libs[@]}"; do
args+=("-l$lib");
done
full_command=("$command" "${args[@]}")
# prepend the ccache binary if we're using ccache
if [ -n "$SPACK_CCACHE_BINARY" ]; then
case "$lang_flags" in
C|CXX) # ccache only supports C languages
full_command=("${SPACK_CCACHE_BINARY}" "${full_command[@]}")
# workaround for stage being a temp folder
# see #3761#issuecomment-294352232
export CCACHE_NOHASHDIR=yes
;;
esac
fi
Restore cc: package search paths come before dependency paths (#4692) Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes. This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end. `<includes> <other_args> <library dirs> <rpaths>` This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build. This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order of two `-L` args affects the search path, but we search for all libraries on the command line using the same search path). We preserve the following: 1. Any system directory in the paths will be listed last. 2. The root package's include/library/RPATH flags come before flags of the same type for any dependency. 3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last) 4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
2018-08-05 12:30:17 +08:00
# dump the full command if the caller supplies SPACK_TEST_COMMAND=dump-args
2016-04-05 03:38:21 +08:00
if [[ $SPACK_TEST_COMMAND == dump-args ]]; then
echo "${full_command[@]}"
exit
elif [[ -n $SPACK_TEST_COMMAND ]]; then
die "ERROR: Unknown test command"
fi
#
# Write the input and output commands to debug logs if it's asked for.
#
2016-04-05 03:38:21 +08:00
if [[ $SPACK_DEBUG == TRUE ]]; then
input_log="$SPACK_DEBUG_LOG_DIR/spack-cc-$SPACK_DEBUG_LOG_ID.in.log"
output_log="$SPACK_DEBUG_LOG_DIR/spack-cc-$SPACK_DEBUG_LOG_ID.out.log"
echo "[$mode] $command $input_command" >> "$input_log"
echo "[$mode] ${full_command[*]}" >> "$output_log"
fi
exec "${full_command[@]}"