Fix link order in cosmopolitan.a

It turned out that the linker was doing the wrong with the amalgamation
library concerning weak stubs. A regression test has been added and new
binaries have been uploaded to https://justine.lol/cosmopolitan/

Ideally this should be fixed by building a tool that turns multiple .a
files into a single .a file with deduplication. As a workaround for now
the cosmopolitan.a build is restructured to not include LIBC_STUBS which
meant technical debt needed to be paid off where non-stub interfaces
were moved to LIBC_INTRIN and LIBC_NEXGEN32E.

Thank @PerfectProductions in #31 for the report!
This commit is contained in:
Justine Tunney 2021-01-16 12:05:41 -08:00
parent 04f1d89f84
commit 9f68d6eee9
121 changed files with 302 additions and 381 deletions

View file

@ -17,14 +17,10 @@ APE_LIB_A_SRCS = \
APE_LIB_A_OBJS = \
$(APE_LIB_A_SRCS_S:%.S=o/$(MODE)/%.o) \
$(APE_LIB_A_SRCS_C:%.c=o/$(MODE)/%.o) \
$(APE_LIB_A_SRCS:%=o/$(MODE)/%.zip.o) \
o/$(MODE)/ape/ape.lds.zip.o \
o/$(MODE)/ape/ape.S.zip.o \
o/$(MODE)/NOTICE.zip.o
$(APE_LIB_A_SRCS_C:%.c=o/$(MODE)/%.o)
APE_LIB_A_CHECKS = $(APE_LIB_A_HDRS:%=o/$(MODE)/%.ok)
APE_LIB_A_DIRECTDEPS = LIBC_STR LIBC_STUBS
APE_LIB_A_DIRECTDEPS = LIBC_STR LIBC_NEXGEN32E LIBC_STUBS
APE_LIB_A_DEPS = $(call uniq,$(foreach x,$(APE_LIB_A_DIRECTDEPS),$($(x))))
$(APE_LIB_A): ape/lib/ $(APE_LIB_A).pkg $(APE_LIB_A_OBJS)