powerpc/mpc55xxevb build failure

[Joel Sherrill]

On 11/23/2013 9:20 AM, Gedare Bloom wrote:
> In fsdosfsname01 I get the following failure during linking
> 
> link command:
> powerpc-rtems4.11-gcc -B../../../../../mpc5566evb/lib/ -specs
> bsp_specs -qrtems -mcpu=8540 -meabi -msdata -fno-common -msoft-float
> -D__ppc_generic -mstrict-align -O2 -g -fno-keep-inline-functions -Wall
> -Wmissing-prototypes -Wimplicit-function-declaration
> -Wstrict-prototypes -Wnested-externs    -mcpu=8540 -meabi -msdata
> -fno-common -msoft-float -D__ppc_generic -mstrict-align   -o
> fsdosfsname01.exe init.o ramdisk_support.o
> 
> failure:
> linkcmds.base:344 cannot move location counter backwards (from
> 000000004002c5e4 to 0000000040020000)
> collect2: error: ld returned 1 exit status
> 
> This is with HEAD (with the latest patch applied to this bsp) and
> current RSB-built tool set.

This looks like a memory region is full and this test won't fit.

>From the linkcmds for this BSP variant.

MEMORY {
        ROM : ORIGIN = 0x0, LENGTH = 3M
        RAM : ORIGIN = 0x40000000, LENGTH = 128K
        RAM_EXT : ORIGIN = 0x20000000, LENGTH = 512K
        NOCACHE : ORIGIN = 0x0, LENGTH = 0
}

It looks like there is nearly 64k too much data.

Moving to per function/variable segments should help this but
whether it will make this test fit, who knows.

My ruling is that this test is too large for this BSP variant.
There is some ifdef magic to handle this if this is the case.

Sebastian should rule on this since I think this is his BSP
family.

> -Gedare
> _______________________________________________
> rtems-devel mailing list
> rtems-devel at rtems.org
> http://www.rtems.org/mailman/listinfo/rtems-devel
> 



-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel.sherrill at OARcorp.com        On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
Support Available                (256) 722-9985