The compat layer on sparc64 was not filling in the nanosecond
fields in properly for 32-bit compat tasks. This caused things
like the test-utime.c test to fail in the libc sources.
A problem still remains for native 64-bit binaries. Like Alpha
the normal stat structure doesn't have the nanosecond fields
so I have to add in the stat64 syscall entry points in the 64-bit
syscall table then add the necessary libc magic.
Signed-off-by: David S. Miller <[EMAIL PROTECTED]>
===== arch/sparc64/kernel/sys_sparc32.c 1.118 vs edited =====
--- 1.118/arch/sparc64/kernel/sys_sparc32.c 2005-03-21 09:56:06 -08:00
+++ edited/arch/sparc64/kernel/sys_sparc32.c 2005-04-05 12:38:54 -07:00
@@ -352,11 +352,11 @@
err |= put_user(old_encode_dev(stat->rdev), &statbuf->st_rdev);
err |= put_user(stat->size, &statbuf->st_size);
err |= put_user(stat->atime.tv_sec, &statbuf->st_atime);
- err |= put_user(0, &statbuf->__unused1);
+ err |= put_user(stat->atime.tv_nsec, &statbuf->st_atime_nsec);
err |= put_user(stat->mtime.tv_sec, &statbuf->st_mtime);
- err |= put_user(0, &statbuf->__unused2);
+ err |= put_user(stat->mtime.tv_nsec, &statbuf->st_mtime_nsec);
err |= put_user(stat->ctime.tv_sec, &statbuf->st_ctime);
- err |= put_user(0, &statbuf->__unused3);
+ err |= put_user(stat->ctime.tv_nsec, &statbuf->st_ctime_nsec);
err |= put_user(stat->blksize, &statbuf->st_blksize);
err |= put_user(stat->blocks, &statbuf->st_blocks);
err |= put_user(0, &statbuf->__unused4[0]);
===== include/asm-sparc64/compat.h 1.19 vs edited =====
--- 1.19/include/asm-sparc64/compat.h 2005-02-17 21:53:03 -08:00
+++ edited/include/asm-sparc64/compat.h 2005-04-05 12:37:58 -07:00
@@ -51,11 +51,11 @@
compat_dev_t st_rdev;
compat_off_t st_size;
compat_time_t st_atime;
- u32 __unused1;
+ u32 st_atime_nsec;
compat_time_t st_mtime;
- u32 __unused2;
+ u32 st_mtime_nsec;
compat_time_t st_ctime;
- u32 __unused3;
+ u32 st_ctime_nsec;
compat_off_t st_blksize;
compat_off_t st_blocks;
u32 __unused4[2];
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
