Make BusyBox fails
 

Hi all,
I am trying to port BusyBox to a TI - TMS board and i follow these steps:
1. I make menuconfig and set using cross compiler option in Build options to y
2. Set the cross compiler prefix properly

then I save the .config.
then export LD_LIBRARY_PATH=/path/to/my/uClibc/lib

Now i do make and woah .. its error!! :confused:
-----------------------------------------------
Command-line error #573: invalid option: 17
1 fatal error detected in this compilation.
Compilation terminated.
/bin/sh: line 0: printf: failure: invalid number
Command-line error #573: invalid option: 17
1 fatal error detected in this compilation.
Compilation terminated.
/bin/sh: line 0: printf: failure: invalid number
------------------------------------------------

and the make blocks. Need to do ctrl-C, to gain control back over shell.

I did run a strace :
--------------------------------------------------
open("Makefile", O_RDONLY|O_LARGEFILE) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=16959, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb7fda000
read(3, "# Makefile for busybox\n#\n# Copyr"..., 4096) = 4096
stat64("/tmp/busybox-1.2.2", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
open("/tmp/busybox-1.2.2/Rules.mak", O_RDONLY|O_LARGEFILE) = 4
fstat64(4, {st_mode=S_IFREG|0644, st_size=15079, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb7fd9000
read(4, "# Rules.make for busybox\n#\n# Cop"..., 4096) = 4096
open("/tmp/busybox-1.2.2/.config", O_RDONLY|O_LARGEFILE) = 5
fstat64(5, {st_mode=S_IFREG|0664, st_size=18185, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb7fd8000
read(5, "#\n# Automatically generated make"..., 4096) = 4096
read(5, " is not set\n# CONFIG_FEATURE_LS_"..., 4096) = 4096
read(5, "FIG_FEATURE_EXTRA_QUIET is not s"..., 4096) = 4096
read(5, "NFIG_FEATURE_HDPARM_HDIO_GETSET_"..., 4096) = 4096
read(5, "DOF_OMIT is not set\n# CONFIG_PS "..., 4096) = 1801
read(5, "", 4096) = 0
close(5) = 0
munmap(0xb7fd8000, 4096) = 0
pipe([5, 6]) = 0
vfork() = 11169
--- SIGCHLD (Child exited) @ 0 (0) ---
sigreturn() = ? (mask now [])
close(6) = 0
read(5, "2006.11.17-06:11+0000\n", 200) = 22
read(5, "", 178) = 0
close(5) = 0
wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 0}], 0, NULL) = 11169
pipe([5, 6]) = 0
vfork() = 11170
close(6) = 0
read(5, Command-line error #573: invalid option: 17
1 fatal error detected in this compilation.
Compilation terminated.
">> Compilation failure\n", 200) = 23
read(5, "", 177) = 0
--- SIGCHLD (Child exited) @ 0 (0) ---
sigreturn() = ? (mask now [])
close(5) = 0
wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 0}], 0, NULL) = 11170
pipe([5, 6]) = 0
vfork() = 11184
close(6) = 0
read(5, /bin/sh: line 0: printf: failure: invalid number
"", 200) = 0
--- SIGCHLD (Child exited) @ 0 (0) ---
sigreturn() = ? (mask now [])
close(5) = 0
wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 1}], 0, NULL) = 11184
pipe([5, 6]) = 0
vfork() = 11185
close(6) = 0
read(5, Command-line error #573: invalid option: 17
1 fatal error detected in this compilation.
Compilation terminated.
">> Compilation failure\n", 200) = 23
read(5, "", 177) = 0
--- SIGCHLD (Child exited) @ 0 (0) ---
sigreturn() = ? (mask now [])
close(5) = 0
wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 0}], 0, NULL) = 11185
pipe([5, 6]) = 0
vfork() = 11199
close(6) = 0
read(5, /bin/sh: line 0: printf: failure: invalid number
"", 200) = 0
--- SIGCHLD (Child exited) @ 0 (0) ---
sigreturn() = ? (mask now [])
close(5) = 0
wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 1}], 0, NULL) = 11199
pipe([5, 6]) = 0
vfork() = 11200
close(6) = 0
read(5,
--------------------------------------------------
The last read we see blocks.
Can anyone tell me why ?! Got my heads cracking over it for no results :scratch:

Did you check out if http://buildroot.uclibc.org/ isn't something for you? If not then I'd first check http://bugs.busybox.net and if no solution is there post it on http://busybox.net/cgi-bin/mailman/listinfo/busybox. BTW, I don't know exactly what logging these syscalls should prove or lead to, I'd be more interested in a verbose "make" instead.