The variable giving the name of the temporary password file was
inconsistently named "ptmp" and "etc_ptmp". This commit changes
it to "passwd_tmp" for consistency and to match OpenBSD's name
for the variable.
- Add some more x86 instructions to emulate,
- emulate VGA read mode 0,
- emulate VGA write mode 1,
- minor cleanup.
Protel's Easytrax, a free PCB layout program, almost runs now; there are
still some problems with the keyboard emulation, but the graphics are fine
(albeit a bit slow).
The current resource buffer returned from an interrupt link device
in the case where there are no interrupts routed for it does not
contain enough space to use it to route an interrupt. In the case
where we need to route an interrupt, throw away the returned buffer
and create a new one containing the interrupt we want.
- Prevent the ACPI code from being loaded as a module other than at
boot time. Loading as a module once the system is up and running
doesn't make any sense.
- Fix acpi_FindIndexedResource (it would only check the first resource),
changes the calling interface.
- Add a new helper function (acpi_AppendBufferResource) to help building
buffers containing resources.
- Remove the beer-ware license (reqested by phk)
- Reorganise so that the PIIX4 workaround code is kept together, and
switch the workaround function via the timecounter struct, saving
a compare in the read-timecounter codepath. Also indicate that
the workaround is active by changing the timecounter hardware string.
Update to version 3.36.
I had to hack print.c due to some new bits that only compile on NetBSD.
I could not figure what the analogous FreeBSD bits are.
Redo how we manage SCSI device settings- have a 3rd flags (nvram) that records
either what's in NVRAM or what the safe defaults would be if we lack NVRAM.
Then we rename cur_XXXX to actv_XXXX (these are the currently active settings)
and the dev_XXX settings to goal_XXXX (these are the settings which we want
cur_XXXX to converge to).
This probably isn't entirely final as yet- but it's a lot closer to now
being what it should be, including allowing camcontrol to actually set
specific settings.
Redo how we manage SCSI device settings- have a 3rd flags (nvram) that records
either what's in NVRAM or what the safe defaults would be if we lack NVRAM.
Then we rename cur_XXXX to actv_XXXX (these are the currently active settings)
and the dev_XXX settings to goal_XXXX (these are the settings which we want
cur_XXXX to converge to).
Redo how we manage SCSI device settings- have a 3rd flags (nvram) that records
either what's in NVRAM or what the safe defaults would be if we lack NVRAM.
Then we rename cur_XXXX to actv_XXXX (these are the currently active settings)
and the dev_XXX settings to goal_XXXX (these are the settings which we want
cur_XXXX to converge to).
Kill the command (don't rerun it) if we had an AUTOSENSE failure.
If we had an AUTOSENSE failure, we don't know what SENSE DATA
we had for a CHECK CONDITION. It's far better to assume failure
in this case.
(Forced commit- last commit was inadvertant in that it missed the comments)
Move TARGCTLIOALLOCUNIT to OTARGCTLIOALLOCUNIT, TARGCTLIOFREEUNIT
to OTARGCTLIOFREEUNIT and redefine old associated structure to be
old_ioc_alloc_unit- deprecation but preservation of binaries.
Add new structure for same- but this one contains a pointer to
user defined INQUIRY data so you can define what the target
device looks like to the outside world.
1. If we get frozen, unfreeze for disable disconnects.
2. Put CAM_DIS_DISCONNECT commands at the head of the work queue
(we have a target still connected and we can't run anything else
until this command completes).
If we had an error sending the last CTIO, unfreeze the queue anyway.
When booted -v (eg bootverbose is non-zero), have pccard report what
resources it is attempting to assign to a child object. This should
help people track down mysterious resource allocation problems more
easily.
# Unfortunately, it is harder to do the conflict check and report which
# resource failed if the driver itself doesn't.
Two recent commits in sys/ufs/ufs interacted badly with ext2fs
because it shares ufs code. In ufs_fhtovp(), the test on i_effnlink
is invalid because ext2fs does not maintain this field. In ufs_close(),
i_effnlink is also tested, to determines whether or not to call
vn_start_write(). The ufs_fhtovp issue breaks NFS exporting of
ext2fs filesystems; I believe the other is harmless.
Fix both cases by checking um_i_effnlink_valid in the ufsmount
struct, and use i_nlink if necessary.
Disallow ATAPI CD transfers that are not a multiple of the device block
size (previously, the transfer size would be rounded up to a multiple of
the block size, which would overflow the buffer).
This fixes panics when doing things like trying to mount audio CD's.
Avoid any chance of being misunderstood as having libelled developers
or developers' vendors without compromising the importance of warning
against bad practice.
This is the traditional BSD libmp interface implemented in terms of
the OpenSSL BIGNUM interface. It is provided for compatibility only
and should not be used in new code.
Rename the GLOB_MAXPATH flag of glob(3) to GLOB_LIMIT to be compatible
with NetBSD and OpenBSD. glob(3) will now return GLOB_NOSPACE with
errno set to 0 instead of GLOB_LIMIT when we match more than `gl_matchc'
patterns. GLOB_MAXPATH has been left as an alias of GLOB_LIMIT to
maintain backwards compatibility.
se [Sat, 28 Jul 2001 21:34:56 +0000 (21:34 +0000)]
Include value of command line argument that causes an error message or
warning in said message, since fetch may be run from a makefile or script
which does not print the command line.
Permit direct swapping to NFS regular files using swapon(2). We
already allow this for NFS swap configured via BOOTP, so it is
known to work fine.
For many diskless configurations is is more flexible to have the
client set up swapping itself; it can recreate a sparse swap file
to save on server space for example, and it works with a non-NFS
root filesystem such as an in-kernel filesystem image.
However, there's still a bug in login.c
because you copy the environment *before* the call to pam_open_session,
which won't set the necessary variables set by /usr/ports/security/pam_ssh.
I didn't do this when I merged the delta to RELENG_4 because I thought
&merged; didn't apply to contributed software since there is one entry
per application which gets updated with the new version number, as
opposed to all the other programs, which get one entry per update.
However, the previous commit removed &merged; from the IPFilter entry,
so perhaps I just didn't look long enough when I did the tcpdump
merge.
Stable requires machine/clock.h to quiet warnings. It isn't
strictly necessary on current, but having it in here makes the diffs with
stable smaller and doesn't hurt anything except for phk's redundant include
finder.
Remove s_strl*(). I am not sure what was thought they accomplished.
When reading the code I had to stop, say "ok, what does *these*
modifications of strl*() do? Pull out grep. Oh, not in add/, maybe above
in ../lib/? Yep. So what do they do? Comments above them are misleading,
guess I'll have to read the code. Oh, they just test strl* against the
size and return the result of the test. Now I can continue to read the
code I was.
The uses of s_strl*() then test that result and errx()'s.
Lets think about the "optimized" code I am removing:
In general the compiler pushes the three args to strl* onto the stack and calls
s_strl*. s_strl* has to indirectly access 3 args from the stack. Then push
them on the stack a 2nd time for the real strl* call. s_strl* then pops the
return from strl* off the stack; or moves it from the register it was returned
in, to the register where tests can happen. s_strl* then pops the three
arguments to strl*. Perform the test, push the result of the test, or move it
from the result register to the return value register. The caller to s_strl*
now has to either pop the return value of s_strl* or move it from the return
value register to the test register. The caller then pops the three args to
s_strl* off the stack (the same args that s_strl* itself had to pop off after
the real call to strl*). The s_strl* caller then performs a simular test to
what has already been done, and conditionally jumps. By doing things this way, we've given the compiler optimizer less to work with.
Also, please don't forget the that call to s_strl* has possibly jumped to code
not in the cache due to being far away from the calling code, thus causing a
pipeline stall.
So where is the "optimization" from s_strl*?
It isn't code clarity.
It isn't code execution speed. It isn't code size either.
Pacify users who get all bent out of shape when they see the "xl%d: command
never completed" message. The RX reset takes longer complete than it
used to, a lot longer in fact than xl_wait() is prepared to wait.
When we do the RX reset in xl_reset(), this cases xl_wait() to time out
and whine. We wait a little extra time now after the RX reset, which
should silence the warning.
Thanks to obrien for finally getting me a box with a NIC that
causes this problem for me to tinker with.
Disable the dirhash sanity check that panics if an unused directory
entry (d_ino == 0) is found in a position that is not the start of
a DIRBLKSIZ block.
While such entries cannot occur normally (ufs always extends the
previous entry to cover the free space instead), they do not cause
problems and fsck does not fix them, so panicking is bad.