On Thu, 21 Feb 2019 at 09:04, Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
>
> Hi Shaokun,
>
> On Thu, Feb 21, 2019 at 1:45 AM Zhangshaokun <zhangshaokun@xxxxxxxxxxxxx> wrote:
> > On 2019/2/20 18:05, Ard Biesheuvel wrote:
> > > On Wed, 20 Feb 2019 at 10:58, Jarkko Sakkinen
> > > <jarkko.sakkinen@xxxxxxxxxxxxxxx> wrote:
> > >>
> > >> On Wed, Feb 20, 2019 at 11:52:52AM +0200, Jarkko Sakkinen wrote:
> > >>> On Wed, Feb 20, 2019 at 05:11:15PM +0800, Zhangshaokun wrote:
> > >>>> There is a compiler failure on arm64 platform, as follow:
> > >>>>
> > >>>> AS arch/arm64/kvm/hyp.o
> > >>>> CC kernel/trace/ring_buffer.o
> > >>>> In file included from security/integrity/ima/ima_fs.c:30:0:
> > >>>> security/integrity/ima/ima.h:176:7: error: redeclaration of enumerator ‘NONE’
> > >>>> hook(NONE) \
> > >>>> ^
> > >>>> security/integrity/ima/ima.h:188:34: note: in definition of macro ‘__ima_hook_enumify’
> > >>>> #define __ima_hook_enumify(ENUM) ENUM,
> > >>>> ^
> > >>>> security/integrity/ima/ima.h:191:2: note: in expansion of macro ‘__ima_hooks’
> > >>>> __ima_hooks(__ima_hook_enumify)
> > >>>> ^
> > >>>> In file included from ./arch/arm64/include/asm/acpi.h:15:0,
> > >>>> from ./include/acpi/acpi_io.h:7,
> > >>>> from ./include/linux/acpi.h:47,
> > >>>> from ./include/linux/tpm.h:26,
> > >>>> from security/integrity/ima/ima.h:25,
> > >>>> from security/integrity/ima/ima_fs.c:30:
> > >>>> ./include/linux/efi.h:1716:2: note: previous definition of ‘NONE’ was here
> > >>>> NONE,
> > >>>> ^
> > >>>> scripts/Makefile.build:276: recipe for target 'security/integrity/ima/ima_fs.o' failed
> > >>>> make[3]: *** [security/integrity/ima/ima_fs.o] Error 1
> > >>>>
> > >>>> I dug it and it is the commit 901615cb916d ("tpm: move tpm_chip definition to include/linux/tpm.h")
> > >>>
> > >>> This results from a new include in tpm.h:
> > >>>
> > >>> #include <linux/acpi.h>
> > >>>
> > >>> Must be fixed either in include/linux/efi.h or security/integrity/ima.h as
> > >>> those files have a name collision. Makes me wonder why neither has taken
> > >>> care of prefixing the constants properly.
> > >>
> > >> Preferably both subsystems should be fixed with proper 'EFI_' and 'IMA_'
> > >> prefixes. Defining a constant named as NONE in a non-generic subsystem
> > >> (e.g. not part of the core data structures of Linux) and especially
> > >> exporting it to include/linux is not too well considered act.
> > >>
> > >
> > > Fixes for this have already been proposed, and should appear in -next shortly
> > >
> > > The EFI one is here
> > > https://mail.google.com/mail/u/0/#label/linux-efi/FMfcgxwBVgrQRjglPkWRqRqVclGgVDnB
> > >
> >
> > Because of no privilege, the website is denied for me. Anyway, it's nice to have been fixed.
>
> Looks like Ard posted a link to a label in his personal gmail mailbox?
>
Silly me.
https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/commit/?h=efi/core&id=5c418dc789a3898717ebf2caa5716ba91a7150b2
