On 2/2/23 10:21, David Marchand wrote: > On Wed, Feb 1, 2023 at 5:54 PM David Marchand <[email protected]> > wrote: >> >> rte_vect.h pulls some AVX512 instrinsics headers added in GCC 12 [1] >> trigger a lot of warnings: >> >> libtool: compile: env "REAL_CC=ccache gcc" "CHECK=sparse -Wsparse-error >> -I ../include/sparse -I ../include -m64 -I /usr/local/include >> " cgcc -target=x86_64 -target=host_os_specs -D__MMX__=1 >> -D__MMX_WITH_SSE__=1 -D__SSE2_MATH__=1 -D__SSE_MATH__=1 >> -D__SSE__=1 -D__SSE2__=1 -DHAVE_CONFIG_H -I. -I.. -I ../include >> -I ./include -I ../lib -I ./lib -Wstrict-prototypes -Wall >> -Wextra -Wno-sign-compare -Wpointer-arith -Wformat >> -Wformat-security -Wswitch-enum -Wunused-parameter >> -Wbad-function-cast -Wcast-align -Wstrict-prototypes >> -Wold-style-definition -Wmissing-prototypes >> -Wmissing-field-initializers -fno-strict-aliasing >> -Wswitch-bool -Wlogical-not-parentheses -Wsizeof-array-argument >> -Wbool-compare -Wshift-negative-value -Wduplicated-cond -Wshadow >> -Wmultistatement-macros -Wcast-align=strict -mssse3 >> -I/home/dmarchan/git/pub/dpdk.org/22.11/install/include >> -include rte_config.h -I/usr/local/include -Werror >> -D_FILE_OFFSET_BITS=64 -g -O2 -MT lib/bfd.lo -MD -MP >> -MF lib/.deps/bfd.Tpo -c ../lib/bfd.c -o lib/bfd.o >> ../lib/bfd.c: note: in included file (through >> /usr/lib/gcc/x86_64-redhat-linux/12//include/immintrin.h, >> /usr/lib/gcc/x86_64-redhat-linux/12//include/x86intrin.h, ...): >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:38:9: >> error: '_Float16' has implicit type >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:38:18: >> error: Expected ; at end of declaration >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:38:18: >> error: got __v8hf >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:62:41: >> error: Expected ; at end of statement >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:62:41: >> error: got { >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:420:32: >> error: Expected ) in expression >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:420:32: >> error: got __A >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:2271:61: >> error: Expected ) in function call >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:2271:61: >> error: got __A >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:2279:61: >> error: Expected ) in function call >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:2279:61: >> error: got __A >> /usr/lib/gcc/x86_64-redhat-linux/12//include/avx512fp16intrin.h:2328:50: >> error: Expected ) in function call >> [...] >> >> Besides, the list of headers by rte_memcpy.h is now out of sync with DPDK. >> OVS should not have to care about this in any case: OVS takes care to >> include the right headers in its sources, and DPDK now checks that its >> exported headers are self-sufficient. >> > > "La nuit porte conseil", and this argument of mine is probably wrong: > some DPDK headers may rely on rte_memcpy.h includes. > I still think it is worth removing those includes.
Should we keep the stdint and add stddef, so this header is self-sufficient? It is using uint8_t and size_t. > > >> There should be no side effect in removing those headers inclusions from >> the rte_memcpy.h fake header. >> >> 1: https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=a68412117fa4 > > I looked a bit more, and it seems that sparse is, at least partially, > at fault here. > The _Float16 type support is missing in sparse. > > _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
