Messages in this thread | | | From | Richard Weinberger <> | Date | Wed, 7 Oct 2020 14:08:41 +0200 | Subject | Re: Build regressions/improvements in v5.9-rc8 |
| |
On Wed, Oct 7, 2020 at 1:51 PM Paul Bolle <pebolle@tiscali.nl> wrote: > > [Added Richard and Anton.] > > Andy Shevchenko schreef op ma 05-10-2020 om 15:58 [+0300]: > > On Mon, Oct 05, 2020 at 11:35:33AM +0200, Geert Uytterhoeven wrote: > > > On Mon, Oct 5, 2020 at 11:33 AM Geert Uytterhoeven <geert@linux-m68k.org> wrote: > > > > JFYI, when comparing v5.9-rc8[1] to v5.9-rc7[3], the summaries are: > > > > - build errors: +3/-6 > > > > Thanks for the report! > > > > > + /kisskb/src/drivers/gpio/gpiolib-cdev.c: error: implicit > > > declaration of function 'in_ia32_syscall' > > > [-Werror=implicit-function-declaration]: => 430:6 > > > + /kisskb/src/drivers/gpio/gpiolib-cdev.c: error: unknown type name > > > 'compat_u64': => 432:4 > > > > > > x86_64/um-all{mod,yes}config > > > > I guess the quick fix is to disable that code for UML, I don't know how IOCTLs > > are working in UML in cases when host - guest - guest app either from: > > 1. x86_64 - x86_64 - i386; > > 2. x86_64 - i386 - i386. > > I ran into this build error too. > > Perhaps the UML maintainers have an idea what to do here. The commit that > triggers this error is 5ad284ab3a01 ("gpiolib: Fix line event handling in > syscall compatible mode").
UML has no ia32 emulation and therefore no in_ia32_syscall(). Maybe you can check for CONFIG_IA32_EMULATION too?
-- Thanks, //richard
| |