>>> On 11.08.16 at 16:01, <arnd@xxxxxxxx> wrote: > On Thursday, August 11, 2016 6:51:33 AM CEST Jan Beulich wrote: >> >>> On 11.08.16 at 14:39, <arnd@xxxxxxxx> wrote: >> > A previous patch added the --no-wchar-size-warning to the Makefile to >> > avoid this harmless warning: >> > >> > arm-linux-gnueabi-ld: warning: drivers/xen/efi.o uses 2-byte wchar_t yet the >> > output is to use 4-byte wchar_t; use of wchar_t values across objects may >> > fail >> > >> > Changing kbuild to use thin archives instead of recursive linking >> > unfortunately brings the same warning back during the final link. >> > >> > This time, we remove the -fshort-wchar flag that originally caused >> > the warning, hopefully fixing the problem for good. I don't see >> > any reason for having the flag in the first place, as the Xen code >> > does not use wchar_t at all. >> >> It uses efi_char16_t, and by dropping -fshort-wchar you'd open >> up a trap for anyone to fall into who were to add wide string >> literals to that same file. EFI using 16-bit characters requires >> code interfacing with EFI to do so too. > > I don't understand. How is this different from other source files > that use efi_char16_t or the wchar_t definition from include/linux/nls.h? Perhaps they didn't run into the issue yet? > As far as I can tell, they all use 16-bit characters, but none of the > others sets the flag. Maybe we should just always build with -fshort-wchar > from the top-level Makefile? Quite possible, unless elsewhere in the tree 4-byte characters are needed. Jan -- To unsubscribe from this list: send the line "unsubscribe linux-arch" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html