Hi Wolfgang,
On Mon, Mar 19, 2012 at 1:34 AM, Wolfgang Denk wd@denx.de wrote:
Dear Simon Glass,
In message 1329177081-29496-13-git-send-email-sjg@chromium.org you wrote:
These calls should not be made directly any more, since bootstage will call the show_boot_...() functions as needed.
Signed-off-by: Simon Glass sjg@chromium.org
Changes in v4:
- Adjust bootstage.h header to fix build warnings (bisect problem)
Actually not all bisect issues have been fixed. Commit 5dc8871 bootstage: Convert progress numbers 1-9 into enums breaks build with this error:
image.c: In function 'boot_get_ramdisk': image.c:898: error: 'BOOTSTAGE_ID_LOAD_RAMDISK' undeclared (first use in this function) image.c:898: error: (Each undeclared identifier is reported only once image.c:898: error: for each function it appears in.)
Yes you are right I'm afraid.
Also, the v4 12/15 commit breaks the ETX094 board when building with ELDK 4.2 due to grown code size:
Before (v2011.12-511-g5ff5539)
text data bss dec hex filename 17752 1026 2585 21363 5373 /work/wd/tmp-ppc/net/libnet.o
After (v2011.12-512-g77060e):
text data bss dec hex filename 17788 1026 2585 21399 5397 /work/wd/tmp-ppc/net/libnet.o
Yes, the increase is small, but the linker map for this board is carefully crafted ;-)
I think I can fix this, though.
Oh dear. The problem is that for my toolchain both builds fails as the code is too large. It doesn't show up as a difference. I tried with ELDK 4.2 - the overflow seems to be only 12 bytes!
Can you fix it by adjusting the bootstage code or is something else required.
Regards, Simon
Best regards,
Wolfgang Denk
-- DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de Wenn das dann in die Hose geht, nehme ich es auf meine Kappe. -- Rudi Völler, 15. Nov 2003