assigned by RISCOS Ltd (and were assigned by Acorn) to be unique across all
manufacturers. This means that associating each one with a manufacturer (and
checking the manufacturer when attaching) is bogus. Thus, we don't do that
any more.
This should have the pleasant side-effect of getting APDL IDE interfaces
working, since they're just ICS ones with a different manufacturer ID.
into kernel_object where this was missing.
(important here because VM_MIN_KERNEL_ADDRESS != 0)
-add some diagnostics
-eliminate some differences to other Utah derived pmaps
into kernel_object where this was missing.
This is a no-op on ports where VM_MIN_KERNEL_ADDRESS==0, ie all but
cesfic.
Confirmed and corrected by Chuck Silvers.
int); this is consistent with RFC2553, the current draft-ietf-ipngwg-
rfc2553bis-05 and the majority of deployed implementations.
Also, there is an effort under way to have this changed back to int in
XBD6/XSH6, XNS, and POSIX-2001 TC1.
successfully emulates a few test program that use poll semaphores,
including the attach-to-file-descriptor-and-select feature.
There are a few issues:
1) at least one ioctl need to set retval. We handle this in irix_sys_ioctl()
by replacing the data argument by a pointer to a strucutre in the stackgap
that carries the real data and retval. The underlying ioctl methods can
therefore retreive both data and retval.
2) usemaclone is a cloning device: each time it is open, it creates a new
context, and ioctl operation on each open file descriptor will lead to
different behavior. This functionnality is available in NetBSD through the
devvp branch. This first implementation does not use devvp yet, but this
should be done later. Currently, we create a new vnode, and we provide our
own vnode operations. Some operation are applied to the cloned vnode, others
are applied to the original vnode. The v_data field is used to hold a
reference to the original vnode so that we can work on it.
3) at least the setattr vnode operation needs some customisation: IRIX
libc relies on the fact that fchmod on /dev/usema will return 0 in case
of failure.
consider a out-of-range partition letter at the end of the special
device node to be a fatal error; just return NULL and let the caller
fall back to FFS.
This fixes the "mount -u /kern/rootdev /" done by the script installer.
XXX this is still gross, and breaks things like
"mount /my/strange/dev/path/b /mnt". Perhaps it should stat the node
and use the minor number as an index instead?
version into different directories. Which version a given port uses is
controlled by the SOFTFLOAT_BITS make variable. This is set to 64 (which
uses the same code we had before) by default. 32-bit platforms that don't
need extended precision support might get better performance by using 32.
Set the ARM port to use the 32-bit version of SoftFloat, since this is more
than a factor of two faster than the 64-bit version. This should get the
floating-point performance back to what it was in 1.5.
For now, we keep the GENERIC/GENERIC_INDY sets, but make the labels a
bit clearer as to their purposes.
XXX: Still need code to install bootloader into volhdr; need either to
integrate into MI installboot or fix up sgivol (I have changes from Scott
for sgivol, just haven't folded them in yet).
directory; the ramdisk build also relies on the termcap entry here, so this
isn't totally useless. All based on work by Scott Taylor (sgimips at
mrynet.com), with bugs most likely added by me.
