[U-Boot] [RFC] [PATCH 00/39] Rework of the debug() macro

[Marek Vasut]

On Saturday, October 22, 2011 05:23:52 PM Albert ARIBAUD wrote:
> Le 22/10/2011 16:19, Marek Vasut a écrit :
> > On Saturday, October 22, 2011 11:48:43 AM Albert ARIBAUD wrote:
> >> Le 22/10/2011 03:18, Marek Vasut a écrit :
> >>> On Saturday, October 22, 2011 02:35:13 AM Albert ARIBAUD wrote:
> >>>> Hi Marek,
> >>>> 
> >>>> Le 22/10/2011 02:16, Marek Vasut a écrit :
> >>>>> This patch series reworks the debug() and debugX() macro in
> >>>>> include/common.h to be compatible with GCC 4.6.
> >>>>> 
> >>>>> This series needs further discussion:
> >>>>> * Some files suffer size growth with gcc4.2, why ?
> >>>>> 
> >>>>>      - Possibility is that the GCC4.2 is incapable to optimize code
> >>>>>      out so deep
> >>>>> 
> >>>>> * Does this change break any platforms? Compile tested on PowerPC.
> >>>>> 
> >>>>>      - Please run MAKEALL on your platforms to test this.
> >>>> 
> >>>> Any git branch we could pull and rebase on our trees rather than
> >>>> am'ing 39 patches one by one?
> >>> 
> >>> Please try:
> >>> 
> >>> git://git.denx.de/u-boot-marex.git , "debug" branch.
> >>> 
> >>> Cheers
> >> 
> >> A differential 'MAKEALL arm' (ELDK42) show a sharp rise in board not
> >> building properly, from 74 to 141 (out of a total of 248).
> > 
> > This was expected ... will fix.
> 
> Much welcome. :)
> 
> >> Of the 67 boards, many show , essentially due to a lot of printf format
> >> warnings such as:
> >> 
> >> ks8695eth.c: In function 'ks8695_eth_send':
> >> ks8695eth.c:199: warning: format '%x' expects type 'unsigned int', but
> >> argument 4 has type 'volatile void *'
> >> 
> >> (in cm4008, cm41xx), or
> >> 
> >> lcd.c: In function 'lcd_setmem':
> >> lcd.c:446: warning: format '%d' expects type 'int', but argument 2 has
> >> type 'u_long'
> >> lcd.c:446: warning: format '%d' expects type 'int', but argument 3 has
> >> type 'u_long'
> >> 
> >> (in at91sam9261ek_nandflash). Another, different, case is smdk2410:
> >> 
> >> s3c2410_nand.c: In function 's3c2410_hwcontrol':
> >> s3c2410_nand.c:59: error: 'DEBUG' undeclared (first use in this
> >> function) s3c2410_nand.c:59: error: (Each undeclared identifier is
> >> reported only once s3c2410_nand.c:59: error: for each function it
> >> appears in.)
> >> s3c2410_nand.c: In function 's3c2410_dev_ready':
> >> s3c2410_nand.c:86: error: 'DEBUG' undeclared (first use in this
> >> function) s3c2410_nand.c: In function 'board_nand_init':
> >> 
> >> Seems like you're trading some warnings against some others, depending
> >> on the compiler. :(
> > 
> > Not trading, the typechecking wasn't there and now that it is there,
> > these kind of warnings keep popping up. Did you ever try to build those
> > boards with DEBUG enabled?
> 
> Not the whole code; I only enabled debug in very specific areas.

So because noone actually bothered to compile u-boot with DEBUG enabled, we (me 
?) now have literally piles of places to fix. Hm :-(

Cheers
> 
> > Cheers
> 
> Amicalement,