Re: mei file2alias compile error

[Geert Uytterhoeven]

On Thu, May 2, 2013 at 5:23 PM, Dave Hansen <dave.hansen@xxxxxxxxx> wrote:
> On 05/01/2013 11:50 PM, Winkler, Tomas wrote:
>>> On Wed, May 01, 2013 at 09:56:35AM -0700, Dave Hansen wrote:
>>>> Linus's current git isn't compiling for me:
>>>>>   HOSTCC  scripts/mod/file2alias.o
>>>>> /home/davehans/linux.git/scripts/mod/file2alias.c: In function
>>> âdo_mei_entryâ:
>>>>> /home/davehans/linux.git/scripts/mod/file2alias.c:1140:1: error:
>>> âOFF_mei_cl_device_id_nameâ undeclared (first use in this function)
>>>>> /home/davehans/linux.git/scripts/mod/file2alias.c:1140:1: note: each
>>> undeclared identifier is reported only once for each function it appears in
>>>>> /home/davehans/linux.git/scripts/mod/file2alias.c: At top level:
>>>>> /home/davehans/linux.git/scripts/mod/file2alias.c:1146:1: error:
>>> âSIZE_mei_cl_device_idâ undeclared here (not in a function)
>>>>
>>>> I confirmed that this commit: e5354107e is the one causing it for me.
>>>> My .config is here:
>>>>
>>>>     http://sr71.net/~dave/linux/config-3.9-meibad
>>>>
>>>> Note that I don't even have CONFIG_INTEL_MEI enabled with this config.
>>>
>>> I don't have CONFIG_INTEL_MEI set either, and I can't duplicate this.
>>> Samuel and Tomas, can you please track this down?
>>
>> I cannot reproduce it as well. I've tried with v3.9 and also with the latest tip.
>
> I did some more poking around.  I was building with O= to put the
> binaries in another dir.  When I saw the error, one of the first things
> I did was do a "make mrproper" on the binary directory.  But, I never
> did one one the source directory.  Doing that seems to have made this go
> away.  There must have been some build gunk in the source directory
> causing this.

Same here.

Removing "scripts/mod/devicetable-offsets.h" from the source directory fixes it.
Unfortunately I forgot to check the date on the bogus
scripts/mod/devicetable-offsets.h,
so I don't know when it was generated.

Usually, if you build in the source directory, a subsequent build in the output
directory will tell you to run "make mrproper" in the source directory. That
obviously didn't happen.

Note that I did run "make tags" in the source directory. So far that
never caused
problems. Just checked, it does not create scripts/mod/devicetable-offsets.h.

JFYI, putting back scripts/mod/devicetable-offsets.h by copying from a
freshly generated one in the output directory, with the mei lines removed,
resurrects the problem.
Even after touching scripts/mod/devicetable-offsets.c afterwards, it is not
regenerated.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/