Hi Andreas, Thank you for your comments!

On Tue, Nov 22, 2011 at 9:49 AM, Andreas Bießmann andreas.devel@googlemail.com wrote:

Dear Christian,

sorry I haven't followed the discussion so far. I have a little pointer too.

Am 22.11.2011 09:26, schrieb Christian Riesch:

...

This patch avoids build breakage for SPLs that do not support printf.

Signed-off-by: Christian Riesch christian.riesch@omicron.at Cc: Wolfgang Denk wd@denx.de Cc: Tom Rini trini@ti.com Acked-by: Tom Rini trini@ti.com

---

Changes for v3:

• Removed extra white space
• Separated patch from patchset [1]

[1] http://lists.denx.de/pipermail/u-boot/2011-November/110635.html

arch/arm/lib/eabi_compat.c |    2 ++  1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/arch/arm/lib/eabi_compat.c b/arch/arm/lib/eabi_compat.c index eb3e26d..e1b87be 100644 --- a/arch/arm/lib/eabi_compat.c +++ b/arch/arm/lib/eabi_compat.c @@ -13,7 +13,9 @@

int raise (int signum)  { +#if !defined(CONFIG_SPL_BUILD) || defined(CONFIG_SPL_LIBCOMMON_SUPPORT)       printf("raise: Signal # %d caught\n", signum); +#endif

Well, you can do it that way but I guess there are several places where printf() will be needed by some drivers in SPL. Therefore I think it is better to provide printf() for SPL than to ifdef out all the printf's.

So how about providing an empty

#if !defined(CONFIG_SPL_BUILD) || defined(CONFIG_SPL_LIBCOMMON_SUPPORT) #define printf(x) #endif

somewhere in the SPL code?

If others agree I will do something like this. But I don't like this approach. I can imagine that this could be a real PITA when you are debugging an SPL, see a printf in the code, expect an output on console, but this output never appears because printf has been replaced globally in some obscure header file.

Regards, Christian