From mboxrd@z Thu Jan  1 00:00:00 1970
From: Simon Glass <sjg@chromium.org>
Date: Mon, 26 Sep 2011 14:04:49 -0700
Subject: [U-Boot] [RFC PATCH v2 13/20] sandbox: Add OS dependent layer
In-Reply-To: <201109260116.10005.vapier@gentoo.org>
References: <1316794942-24709-1-git-send-email-sjg@chromium.org>
	<1316794942-24709-14-git-send-email-sjg@chromium.org>
	<201109260116.10005.vapier@gentoo.org>
Message-ID: <CAPnjgZ3vEz2STXJ0YtOoo7TjMhnR_=-YGHK_WJbFWaSLEGo3kA@mail.gmail.com>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

Hi Mike,

On Sun, Sep 25, 2011 at 10:16 PM, Mike Frysinger <vapier@gentoo.org> wrote:
> On Friday, September 23, 2011 12:22:15 Simon Glass wrote:
>> ?arch/sandbox/cpu/sandbox/Makefile | ? ?7 ++++-
>> ?arch/sandbox/cpu/sandbox/os.c ? ? | ? 49
>
> you don't need to follow the "SoC" convention. ?arch/sandbox/cpu/*.c should
> work fine ...

OK, will do.

>
>> --- a/arch/sandbox/cpu/sandbox/Makefile
>> +++ b/arch/sandbox/cpu/sandbox/Makefile
>>
>> +# I want to do this, but it doesn't seem to work
>> +CFLAGS_arch/sandbox/cpu/sandbox/os.o += -I/usr/include
>
> seems to work for me ...
> CFLAGS_arch/blackfin/lib/board.o = -ffoo

That is not a -I flag, though. If you try -I I think you will get a
failure. Looking a bit closer it is because the dependency stuff uses
CPPFLAGS which doesn't support per file/dir flags.

>
> if you edit config.mk like so, does that help you figure out what is wrong ?
>
> --- a/config.mk
> +++ b/config.mk
> @@ -280,6 +280,8 @@
> ?$(obj)%.s: ? ? %.S
> ? ? ? ?$(CPP) $(ALL_AFLAGS) -o $@ $<
> ?$(obj)%.o: ? ? %.S
> + ? ? ? echo $(BCURDIR)/$(@F)
> + ? ? ? echo $(ALL_CFLAGS)
> ? ? ? ?$(CC) ?$(ALL_AFLAGS) -o $@ $< -c
> ?$(obj)%.o: ? ? %.c
> ? ? ? ?$(CC) ?$(ALL_CFLAGS) -o $@ $< -c
>
>> --- /dev/null
>> +++ b/arch/sandbox/cpu/sandbox/os.c
>>
>> +int os_open(const char *pathname, int flags)
>> +{
>> + ? ? return open(pathname, flags);
>> +}
>
> i guess since u-boot can't really create files, we don't need to support the
> 3rd arg (mode) ... what about creating file-backed flash devices on the fly ?
> -mike
>

Who would ever want that? :-) Even if we do it's not clear that we
want to set the mode, so would prefer to leave this bit for now.

Regards,
Simon