arm-unknown-linux-gnueabi ABI selection

[Yann E. MORIN]

On Sunday 31 August 2008 17:49:30 Michael Abbott wrote:
> 
> On Sun, 31 Aug 2008, Michael Abbott wrote:
> > > 2. I can't build a good toolchain (sizeof(enum)==1 is a disaster).
> > I may have spotted a discrepancy, with CT_ARCH_ABI_CFLAGS where it looks 
> > as if CT_ARCH_ABI_CFLAG was meant.  
> ...
> > I'll make an update in an hour when the build is complete.
> And ... it still doesn't work.

OK, working on Marc's idea, I managed to have a 'working' EABI toolchain:

# cat test.c
enum test { TEST };
int size() { return sizeof(enum test); }
# ./arm-unknown-linux-gnueabi/bin/arm-unknown-linux-gnueabi-gcc -O2 -S test.c
# cat test.s
        .arch armv5te
        .fpu softvfp
        .eabi_attribute 20, 1
        .eabi_attribute 21, 1
        .eabi_attribute 23, 3
        .eabi_attribute 24, 1
        .eabi_attribute 25, 1
        .eabi_attribute 26, 2
        .eabi_attribute 30, 2
        .eabi_attribute 18, 4
        .file   "test.c"
        .text
        .align  2
        .global size
        .type   size, %function
size:
        @ args = 0, pretend = 0, frame = 0
        @ frame_needed = 0, uses_anonymous_args = 0
        @ link register save eliminated.
        mov     r0, #4
        bx      lr
        .size   size, .-size
        .ident  "GCC: (crosstool-NG-1.2.0+svn_trunk@970M) 4.3.1"
        .section        .note.GNU-stack,"",%progbits
# ./arm-unknown-linux-gnueabi/bin/arm-unknown-linux-gnueabi-gcc -O2 \
  -c test.c -o test.o
# ./arm-unknown-linux-gnueabi/bin/arm-unknown-linux-gnueabi-readelf -h test.o
ELF Header:
  Magic:   7f 45 4c 46 01 01 01 00 00 00 00 00 00 00 00 00
  Class:                             ELF32
  Data:                              2's complement, little endian
  Version:                           1 (current)
  OS/ABI:                            UNIX - System V
  ABI Version:                       0
  Type:                              REL (Relocatable file)
  Machine:                           ARM
  Version:                           0x1
  Entry point address:               0x0
  Start of program headers:          0 (bytes into file)
  Start of section headers:          236 (bytes into file)
  Flags:                             0x4000000, Version4 EABI
  Size of this header:               52 (bytes)
  Size of program headers:           0 (bytes)
  Number of program headers:         0
  Size of section headers:           40 (bytes)
  Number of section headers:         10
  Section header string table index: 7

Seems in the case of EABI, we must entirely leave out the --with-abi and -mabi
flags, as Marc said. So, is the above an EABI-compliant toolchain?

Michael, care to test further the attached patch? If that's OK, then I'll
remove the ABI config option in case EABI is selected.

BTW, Michael, I saw you were using gcc-4.2.4. I'm using 4.3.1, and got the
same results as yours without the patch. I did not test the patch with 4.2.4.

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +0/33 662376056 | Software  Designer | \ / CAMPAIGN     |   ^                |
| --==< ^_^ >==-- `------------.-------:  X  AGAINST      |  /e\  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |  """  conspiracy.  |
`------------------------------^-------^------------------^--------------------'
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ct-ng.ARM-EABI.patch
Type: text/x-diff
Size: 1142 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/crossgcc/attachments/20080831/10797fe9/attachment.bin>
-------------- next part --------------
--
For unsubscribe information see http://sourceware.org/lists.html#faq