On 12/12/19 4:11 PM, Ken Moffat via lfs-dev wrote:
I'm trying to build the current books to look at the state of play
with dropping python2 for my own desktop builds. I know that isn't
going to be achievable at the moment (e.g. gimp-2.10 will require
py2, but gimp-3 will be ok whenever it is ready). But I'm seeing
test failures in libffi-3.3.
6 unexpected failures, all in libffi.bhaible/test-callback.c, two
each for tests 54,55,56 [ -DDGTEST=54 etc ]
Running ../../testsuite/libffi.bhaible/bhaible.exp ...
FAIL: libffi.bhaible/test-callback.c -W -Wall -Wno-psabi -DDGTEST=54
-Wno-unused-variable -Wno-unused-parameter -Wno-unused-but-set-variable
-Wno-uninitialized -O0 -DABI_NUM=FFI_GNUW64 -DABI_ATTR=__MSABI__ execution test
FAIL: libffi.bhaible/test-callback.c -W -Wall -Wno-psabi -DDGTEST=54
-Wno-unused-variable -Wno-unused-parameter -Wno-unused-but-set-variable
-Wno-uninitialized -O2 -DABI_NUM=FFI_GNUW64 -DABI_ATTR=__MSABI__ execution test
FAIL: libffi.bhaible/test-callback.c -W -Wall -Wno-psabi -DDGTEST=55
-Wno-unused-variable -Wno-unused-parameter -Wno-unused-but-set-variable
-Wno-uninitialized -O0 -DABI_NUM=FFI_GNUW64 -DABI_ATTR=__MSABI__ execution test
FAIL: libffi.bhaible/test-callback.c -W -Wall -Wno-psabi -DDGTEST=55
-Wno-unused-variable -Wno-unused-parameter -Wno-unused-but-set-variable
-Wno-uninitialized -O2 -DABI_NUM=FFI_GNUW64 -DABI_ATTR=__MSABI__ execution test
FAIL: libffi.bhaible/test-callback.c -W -Wall -Wno-psabi -DDGTEST=56
-Wno-unused-variable -Wno-unused-parameter -Wno-unused-but-set-variable
-Wno-uninitialized -O0 -DABI_NUM=FFI_GNUW64 -DABI_ATTR=__MSABI__ execution test
FAIL: libffi.bhaible/test-callback.c -W -Wall -Wno-psabi -DDGTEST=56
-Wno-unused-variable -Wno-unused-parameter -Wno-unused-but-set-variable
-Wno-uninitialized -O2 -DABI_NUM=FFI_GNUW64 -DABI_ATTR=__MSABI__ execution test
Running ../../testsuite/libffi.call/call.exp ...
Running ../../testsuite/libffi.closures/closure.exp ...
Running ../../testsuite/libffi.complex/complex.exp ...
Running ../../testsuite/libffi.go/go.exp ...
=== libffi Summary ===
# of expected passes 2278
# of unexpected failures 6
The tests are all in test-callback.c but the details of what exactly
these are doing are beyond me.
I can read the README in libffi.bhaible, that implies that there
should be output files but I can't see any, nor indeed the compiled
programs. Trying to manually run make test-call or make
test-callback says it can't find <ffi.h>.
The ffi.h is in x86_64-pc-linux-gnu/include/ffi.h so I symlinked
that (prefixed ../) to include, but of course that is not where it
is looking. Then I changed test-call*.c to use "ffi.h" which, with
the symlink, does exist, but still no joy:
root in chroot /building/libffi-3.3# make -C testsuite/libffi.bhaible/
test-callback
make: Entering directory '/building/libffi-3.3/testsuite/libffi.bhaible'
gcc -I/include -O2 -Wall -L/lib -Wl,-rpath,/lib -o test-callback
test-callback.c -lffi
test-callback.c:24:10: fatal error: ffi.h: No such file or directory
24 | #include "ffi.h"
| ^~~~~~~
compilation terminated.
make: *** [Makefile:15: test-callback] Error 1
make: Leaving directory '/building/libffi-3.3/testsuite/libffi.bhaible'
My initial thought when I saw failing tests was that I need to make
some more static libs visible (people will recall that I have an
aversion to static libs being picked up silently), and that might
still be true - but like most things related to compsci, this is
beyond me so I'll let these tests fail.
Have you noticed that we have those failures documented?
When I last updated libffi I looked at test-callback.c. The
-DDGTEST=5{4,5,6} variables indicate which test to run. Each is run at
-O0 and -O2. At that point I could not figure out what was failing.
And just documented the failures.
-- Bruce
--
http://lists.linuxfromscratch.org/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
