Andrew's problems with SCSI on some alphas- do not call isp_update
directly to update parameters- just mark them as being ready to
update for the next command- the system would just hang on a READ
CAPACITY for a drive. Really annoying because it wouldn't even timeout
(and it has a timeout) so either the SET PARAMETERS call was nuking
things or the f/w was really dropping the ball.
approved: jkh
Reviewed by: gallatin@freebsd.org
is gone as a define. We just don't support fast posting for anything less
than the 1240/1080/1280/12160 or Fibre Channel cards.
Put in support for CDB's larger than 12 bytes for parallel SCSI (up to 44
bytes are allowed).
Approved: jkh
- Remove all the code intended to deal with experimental
C1010 revisions. This code got useless due to commercial
chip revisions having been fixed.
Fixes:
- Rewrite/rework the WSR condition handling.
Previous drivers snooped on the BUS through the SBDL IO
register and this has been discovered to trigger a spurious
SCSI parity error when WSR had been set by chip and cleared
by SCRIPTS prior to reading SBDL bit [0...7].
On the other hand, the C1010 does not use the SWIDE register
when synchronous data transfers are taking place and
requires a CHMOV (1) WHEN DATA_IN to be performed in order
to move to memory the residual byte when WSR is set and
the residual byte is useful data.
BTW, the new WSR handling by the driver is simpler.
- No longer attempt to read from SCRIPTS the SBDL register.
This is intended to avoid to be victimized again by any other
issue regarding the handling of this register by 8xx chips.
Miscellaneous:
- The driver is now able to handle the WSR + IGN RESIDUE
condition at the end of a DATA IN I/O without need of a
programmed interrupt. It is a minor? optimization.
- A few other minor cosmetic changes.
This driver version fixes notably a permanent SCSI parity
error condition at boot that can be triggerred due to recent
changes in cam_xpt.c between 1.79 and 1.80.
Changes in CAM/XPT are fine, but the new handling of the full
INQUIRY may trigger the driver problem when a target returned
an odd value in the `additionnal length' field of the INQUIRY
response.
The diff against previous driver version is large, but it
consists approximatively in:
- 350 lignes removed and not compiled in previous drivers
(They addressed experimental C1010 revisions)
- 250 lignes added or changed, half being comments or empty
lines.
So, in fact, the real changes are about 120 lines of source.
About 80 lines address SCRIPTS changes and about 40 lines
address C code changes.
Approved by: jkh
New WSR handling reviewed by Pamela Delaney <pam.delaney@lsil.com>
(For back-porting to Linux sym53c8xx driver 1.6x series)
``jail'', and move the set_hostname_allowed sysctl there, as well as
fixing a bug in the sysctl that resulted in jails being over-limited
(preventing them from reading as well as writing the hostname). Also,
correct some formatting issues, courtesy bde :-).
Reviewed by: phk
Approved by: jkh
for 1020/1X80/12160/2X00- for readability. Add in 12160 (Ultra3)
support- but not with PPR just yet. Fix and clarify fetching of
return parameter for getting firmware rev which for the 2200 contains
the connection topology (Private Loop (NL-port), N-port, FL-port,
F-port). Synthesize the connection topology for the 2100 which can
only be Private Loop or FL-port. Handle a couple of new async
mailbox commands which signify connection in Point-to-Point mode
(N-port or F-port) or indicate various toe stubbing getting to same.
Approved: jkh@freebsd.org
- trying for a fit for a PnP configuration from a device
- soaking up resources from a configuration that were not allocated by
the driver
do not attempt to activate them. Only a device driver that is aware of
the nature of the resource and its suitability can be certain that
activating a resource, particularly a memory resource, is a safe thing
to do.
This was prompted by the discovery that many systems report all physical
memory through a PNP0c02 device; activating this resource maps all
physical memory into the kernel's virtual space, either blowing out the
kernel pagetable or in the worst case causing a panic in pmap_mapdev()
if the system has too much physical memory.
Authorised by: jkh
Reviewed by: dfr
won't appear on the screen, and "blinking" and "destructive" cursor
won't appear in the vty for which the text cursor is currently hidden.
Approved by: jkh
Some of IPsec authentication related functions should have
'const' for its 2nd argument, but not now.
But if someone try to use them, and passed const data for
those functions, then much bogus compile warnings will be
generated.
So those funcs prototype should be modified.
Requested by: archie
Approved by: jkh
KAME put INET6 related stuff into sys/netinet6 dir, but IPv6
standard API(RFC2553) require following files to be under sys/netinet.
netinet/ip6.h
netinet/icmp6.h
Now those header files just include each following files.
netinet6/ip6.h
netinet6/icmp6.h
Also KAME has netinet6/in6.h for easy INET6 common defs
sharing between different BSDs, but RFC2553 requires only
netinet/in.h should be included from userland.
So netinet/in.h also includes netinet6/in6.h inside.
To keep apps portability, apps should not directly include
above files from netinet6 dir.
Ideally, all contents of,
netinet6/ip6.h
netinet6/icmp6.h
netinet6/in6.h
should be moved into
netinet/ip6.h
netinet/icmp6.h
netinet/in.h
but to avoid big changes in this stage, add some hack, that
-Put some special macro define into those files under neitnet
-Let files under netinet6 cause error if it is included
from some apps, and, if the specifal macro define is not
defined.
(which should have been defined if files under netinet is
included)
-And let them print an error message which tells the
correct name of the include file to be included.
Also fix apps which includes invalid header files.
Approved by: jkh
Obtained from: KAME project
problem was basically (for offset > 4096):
vtophys(addr) + offset != vtophys(addr + offset)
Also, use TD's with a maximum size of 4k instead of 8kb for OHCI
controllers.
This problem occurs in drivers that use large transfer sizes:
umass, host2host and ethernet with jumbo frames.
or not a process in a jail, with privilege, may set the jail's hostname.
Defaults to 1, which permits this. May be set to 0 by a process with
appropriate privilege outside of jail. Preventing hostname renaming
from within a jail is currently required to make jails manageable, as they
a currently identifiable only by hostname using /proc, which may be
modified without this sysctl being set to 0. This will be documented
in upcoming man commits.
Authorized by: jkh, the ever-patient
for optimizing the unpause operation no-longer exist, and this is much
safer.
When restarting the sequencer, reconstitute the free SCB list on the card.
This deals with a single instruction gap between marking the SCB as free
and actually getting it onto the free list.
Reduce the number of transfer negotiations that occur. In the past, we
renegotiated after every reported check condition status. This ensures
that we catch devices that have unexpectidly reset. In this situation,
the target will always report the check condition before performing a
data-phase. The new behavior is to renegotiate for any check-condition where
the residual matches the orginal data-length of the command (including
0 length transffers). This avoids renegotiations during things like
variable tape block reads, where the check condition is reported only
to indicate the residual of the read.
Revamp the parity error detection logic. We now properly report and
handle injected parity errors in all phases. The old code used to hang
on message-in parity errors.
Correct the reporting of selection timeout errors to the XPT. When
a selection timeout occurs, only the currently selecting command
is flagged with SELTO status instead of aborting all currently active
commands to that target.
Fix flipped arguments in ahc_match_scb and in some of the callers of this
routine. I wish that gcc allowed you to request warnings for enums passed
as ints.
Make ahc_find_msg generically handle all message types.
Work around the target mode data-in wideodd bug in all non-U2 chips.
We can now do sync-wide target mode transfers in target mode across the
hole product line.
Use lastphase exclusively for handling timeouts. The current phase
doesn't take the bus free state into account.
Fix a bug in the timeout handler that could cause corruption of the
disconnected list.
When sending an embedded cdb to a target, ensure that we start on a
quad word boundary in the data-fifo. It seems that unaligned stores
do not work correctly.
chipset to be probed & attached on newer Dell PowerEdge servers, such as
the 2400 and 4400.
Reviewed by: dfr, msmith, jlemon
Tested by: hnokubi@yyy.or.jp (in a previous incantation)
Approved by: jkh
