I'm trying to recompile util-linux source under a sarge
dist and it's failing. Same source as I used only days
ago so I suspect something in an upgrade caused the problem.
I've used source both from upstream and via apt-get. The
only thing that changes is in which .c the error shows
up first. The below is from upstream sourceL
cc -pipe -O2 -mcpu=i486 -fomit-frame-pointer -I../lib -Wall -Wmissing-prototypes
-Wstrict-prototypes -DNCH=1 -D_FILE_OFFSET_BITS=64 -DSBINDIR=\"/sbin\"
-DUSRSBINDIR=\"/usr/sbin\" -DLOGDIR=\"/var/log\" -DVARPATH=\"/var\"
-DLOCALEDIR=\"/usr/share/locale\" -O2 -s blockdev.c -o blockdev
blockdev.c:70: error: parse error before '[' token
blockdev.c:70: error: initializer element is not constant
blockdev.c:70: error: (near initialization for `bdcms[4].ioc')
blockdev.c:70: error: initializer element is not constant
blockdev.c:70: error: (near initialization for `bdcms[4]')
blockdev.c:73: error: parse error before '[' token
blockdev.c:73: error: initializer element is not constant
blockdev.c:73: error: (near initialization for `bdcms[5].ioc')
blockdev.c:73: error: initializer element is not constant
blockdev.c:73: error: (near initialization for `bdcms[5]')
blockdev.c:76: error: initializer element is not constant
blockdev.c:76: error: (near initialization for `bdcms[6]')
blockdev.c:79: error: initializer element is not constant
blockdev.c:79: error: (near initialization for `bdcms[7]')
blockdev.c:82: error: initializer element is not constant
blockdev.c:82: error: (near initialization for `bdcms[8]')
blockdev.c:85: error: initializer element is not constant
blockdev.c:85: error: (near initialization for `bdcms[9]')
blockdev.c:89: error: initializer element is not constant
blockdev.c:89: error: (near initialization for `bdcms[10]')
blockdev.c: In function `report_device':
blockdev.c:331: error: parse error before '[' token
make[1]: *** [blockdev] Error 1
make[1]: Leaving directory `/dma/Sbuilder2/Obj-mourne/util-linux-2.12/disk-utils'
make: *** [all] Error 1
There are actually only 3 errors; the rest are just
caused by them:
blockdev.c:70: error: parse error before '[' token
blockdev.c:73: error: parse error before '[' token
blockdev.c:331: error: parse error before '[' token
I narrowed things down by doing a macro expansion. blockdev.c
line 70 contains this:
#ifdef BLKSSZGET
{ "--getss", "BLKSSZGET", BLKSSZGET, ARGINTG, -1, NULL, N_("get sectorsize") },
#endif
which macro expands to this:
{ "--getbsz", "BLKBSZGET", (((2U) << (((0 +8)+8)+14)) | (((0x12)) << (0 +8)) |
(((112)) << 0) | (((((sizeof(sizeof(int)) == sizeof(sizeof(int)[1]) &&
sizeof(sizeof(int)) < (1 << 14)) ? sizeof(sizeof(int)) :
__invalid_size_argument_for_IOC))) << ((0 +8)+8))), 4, -1, ((void *)0), ("get
blocksize") },
If you look closely you will see sizeoff(int)[1]. That
is the error which occurs in all three lines here (and
also once in get_blocks.c.
Has anyone else run across this? It looks like a problem
in the _IOR and _IOW macros.
I'm not subscribed to the list, but I'll check back via
the web archive.
--
------------------------------------------------------
Dale Amon [EMAIL PROTECTED] +44-7802-188325
International linux systems consultancy
Hardware & software system design, security
and networking, systems programming and Admin
"Have Laptop, Will Travel"
------------------------------------------------------
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
