https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123509
Bug ID: 123509
Summary: Switch int8_t etc. to byte on Solaris
Product: gcc
Version: 16.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: d
Assignee: ibuclaw at gdcproject dot org
Reporter: ro at gcc dot gnu.org
Target Milestone: ---
Target: *-*-solaris2.11
Created attachment 63303
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=63303&action=edit
Proposed patch
int8_t on Solaris is of type char (which is signed), in violation of C99. This
is constantly causing problems, especially for C++. Thus I'm considering
a fix using fixincludes, as detailed in PR libstdc++/113450.
There are only two group of regressions when bootstrapping:
+UNRESOLVED: gdc.test/runnable_cxx/stdint.d compilation failed to produce
executable
+UNRESOLVED: gdc.test/runnable_cxx/stdint.d -shared-libphobos compilation
failed to produce executable
Undefined first referenced
symbol in file
_Z15testCppI8Manglechchch /var/tmp//ccJLlOBa.o
+FAIL: libphobos.phobos/std_complex.d execution test
+FAIL: libphobos.phobos/std_format_package.d execution test
+FAIL: libphobos.phobos/std_format_write.d execution test
+FAIL: libphobos.phobos/std_math_hardware.d execution test
+FAIL: libphobos.phobos/std_typecons.d execution test
This can easily be avoided by switching int8_t etc. from char to byte in
stdint.d, as in the attached patch. This way, all regressions are gone.
I guess such a patch would have to go upstream first?
