2441, and part of 2600. 2436 is left open to remind me that Chris's
comments are probably correct. This file should be built statically and
committed to the tree, not built on a per-kernel basis.
string with 0xFF instead of 0x20. Shorten the compare to 6 bytes to
take care of this.
Fixes PR #2617 by myself, which is based on a private problem report
by Neil Brewitt.
- en_mfix shouldn't touch M_EXT mbufs. change to avoid this [at
the expense of allocating a new cluster mbuf to copy to].
XXX: en_mfix is getting called more often than I hoped it would
(seems to happen when TCP retransmits... we get all sorts of odd
sized, odd lengthed data mbufs, yuck). i may revise the xmit
DMA code to use non-word sized dma.
- don't try and add a PDU trailer (or TBD for that matter) to an M_EXT
mbuf. when M_EXT is set, the data area of the mbuf can be shared
with other mbufs (see m_copym()), and writing to one M_EXT mbuf may
invalidate another M_EXT mbuf (saw this with PDU trailers).
- be sure to validate the length value from the PDU trailer so that it
doesn't throw us out of sync with the chip (fixes yet another source
of the dread "en_service: id mismatch" panic).
- on recv, don't bother allocating an mbuf if mlen is zero since we are
only going to toss the data anyway.
function is called when carrier is lost in lieu of a console message.
This is used by the sparc port to switch media types on Sun4m machines
if the appropriate link flags are set by ifconfig (see following commit
message). Should not affect ports that do not use sc_nocarrier().
old config anymore, all devices should use new config by now.
Add structures to handle DMA devices using new config.
Fixed bug that caused unwanted ubareset's on 11/780.
to ignore inbound data sizes less than the threshold. otherwise,
when connected to a video aal0 input the driver gives us a massive
stream of 56 byte mbufs each with one aal0 cell in it and the
system just can't keep up with it, especially if the socket buffer
size is large [it hangs until you turn off the video source].
fixes: when turning off a vc, try and check the freshest copy of the mode
when seeing if we need to enter the "drain" state. also, don't
panic if we get unexpected rx interrupt on a VCI (instead make sure
the VC is off, print a warning, and move on!).
- change rxso into a void handle ... no need to know details at this low
level.
fix:
- forgot to init "vci" during ENOTHER_DRAIN case of RX DMA interrupt
[could cause a vm_fault on native mode connection close depending
on needing to drain out the VC and also the random value of vci]
- more stat counters (rxmbufout -- mbufs, ttrash -- trashed RBD)
- add swsl_size to keep track of size of software slist
- revise when we call en_service
- fix handling of circular lists (was getting full vs empty confused in
extreem cases...)
- native mode atm: remove sbappend code: can't do it here since
socket buffer is protected only by splsoftnet and that doesn't
mask enintr off. forces us to switch to a two level interrupt
scheme. revise call to atm_input to reflect that.
[1] rxslot "cur" pointer wasn't updated if we copied all data and
didn't have any fill. this could cause a service id mismatch
panic under load.
[2] recv of aal0 forgot to include the cell header in the data length
calc [add cell header length to midwayreg.h]
[3] fix control info "chop" for raw mode and aal0
[4] handle rxsb better for native atm connections
(XXX: still working on this ... not complete and needs revision)
Enable reselections as soon as possible after a disconnect - prevents
losing a reselecting device.
Check for and ignore a spurious interrupt during a DMA input (from the
Mach driver).
Turbochannel as well as the IOASIC. It should now work on the 5000/200.
Removed the "aborts" which could leave the disk trashed when the abort
rebooted the system. Fix the data corruption problem by clearing the
FIFO before starting a data transfer.
all attempts to mode select, even if the attempted mode is supported.
Add the ST_Q_UNIMODAL quirk for the HP Colorado T4000s which exhibits
this behavior. (Someone please lob a tactical nuke in that direction.)
From David Rosenthal <dshr@vitria.com> on netbsd-bugs.
* Change MachEmptyWriteBuffer() to wbflush(). Should use TC mi names tc_mb(),
tc_wmb, tc_syncbus() but I'm not sure which each wbflush() should be.
* Add prototyped forward decl for asc_Dumplog() and add an explicit void
return type.
* Remove unused variables.
alpha. At some point in the future, those headers should be provided
on the pmax, but until then, things are #ifdef'd so that the old
headers and definitions are still used. Pass a bus_chipset_tag_t
to the TC bus attach code and to devices, on the alpha.
to the ring (or silo). This eliminates ring overrun with crtscts.
Also handle deferred register changes earlier, and misc. cleanup.
Give credit to: Bill Studenmund and Ian Dall (Thanks!)
While I'm here, convert driver to use <machine/bus.h> and the new PCI
I/O space interface (the new PCI interrupt interface was already in-use).
Also, correct a "corrected" printf.
bits that tell the driver not to expect parity from those targets.
XXX: For now, it is up to MD code to set this mask. Eventually,
XXX: we should identify targets first (without demanding parity)
XXX: and then use a new "SCSI quirk" to set the right bit.
naming conflicts between bus attachments on ports that can have
multiple instances of the LANCE.
Changed struct ifnet to have a pointer to the softc of the underlying
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_reset)()
to take a struct ifnet *, rather than a unit number.
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_reset)()
to take a struct ifnet *, rather than a unit number.
naming conflicts between bus attachments on ports that can have
multiple instances of the LANCE.
Add a real PCI front-end for PCnet-PCI Ethernet cards.
Changed struct ifnet to have a pointer to the softc of the underlying
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_reset)()
to take a struct ifnet *, rather than a unit number.
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_reset)()
to take a struct ifnet *, rather than a unit number.
naming conflicts between bus attachments on ports that can have
multiple instances of the LANCE.
Break up ISA LANCE driver into ISA and PCI front-ends.
Changed struct ifnet to have a pointer to the softc of the underlying
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_start)()
to take a struct ifnet *, rather than a unit number.
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_start)()
to take a struct ifnet *, rather than a unit number.
naming conflicts between bus attachments on ports that can have
multiple instances of the LANCE.
Changed struct ifnet to have a pointer to the softc of the underlying
device and a printable "external name" (name + unit number), thus eliminating
if_name and if_unit. Updated interface to (*if_watchdog)() and (*if_start)()
to take a struct ifnet *, rather than a unit number.
(and only tested on) Intel Etherexpress PRO. Should work on any i82595 ISA
board which uses an EEPROM for config information; NETBLKID not yet supported.
TODO: change to BUS interface, add multicast support.
Use the i/o handle for the `delay port' exported from the parent
ISA bus, rather than the totally bogus combination of [ioh, 0x84].
Also, fix a bug apparently introduced when the -Wall changes were checked
in which, under most circumstances, would cause a 16-bit WD8013EP to
be mis-identified as an 8-bit WD8003EP.
For the ISA probe, implement a way to keep track of which ISA busses
have been through an ep_isa probe; the static variable that was used
before didn't scale.
In elink_reset(), since some machines may have more than one ISA bus,
implement a way of keeping track of which ISA bus has received an
elink_reset(); The static variable which was used previously didn't
scale.
bus-independent core driver. Tested on all three bus types, including
an isa 3c509 masquerading as an eisa device (use ep* at eisa? slot ? in
your kernel config file to catch this one).
XXX Driver still needs to be converted to <machine/bus.h>
remove their 'integrate' (usually defined to be 'static') keywords.
when lance drivers are split up by attachment, more than one file will
reference the copy/zero functions (i.e. not just the file that pulls in
am7990.c... and eventually inclusion of am7990.c should go away entirely).
chip-independent clock code. 'clock' has been renamed 'mcclock' 'clock'
has been renamed 'mcclock' (since it's a driver for that particular
clock, and since eventually there may be another clock chip driver),
and now attaches via seperate match/attach functions to both TC and ISA.
This removes a whole lot of #ifdefs...
'cflag' for the console. Normally set to TTYDEF_CFLAG, may be
overridden by machine-dependent console attachment code, as necessary.
(Alpha uses it to set cs8 -parenb.) Files including comvar.h now
need to include <sys/termios.h>, because comconscflag is of type
tcflag_t.
tech-kern (?)), to get the IPL_* and IST_* types. pass the machine-dependent
cookie given at attach time on to sub-devices, so they can use it to access
the interrupt functions properly.
the isa_chipset_tag_t type, and to define or prototype:
isa_attach_hook()
isa_intr_establish()
isa_intr_disestablish()
i.e. the machine-dependent functions to be used my MI ISA code.
Remove prototypes for the latter two functions from this file.
the bus is configured, interrupts are handled, etc. machine-dependent
header (machine/eisa/eisa_machdep.h) defines eisa_chipset_tag_t and
eisa_intr_handle_t types, and the following functions:
eisa_attach_hook()
eisa_maxslots()
eisa_intr_map()
eisa_intr_string()
eisa_intr_establish()
eisa_intr_disestablish()
Don't define the interrupt establish/disestablish functions directly
in terms of the ISA functions any more (on some wacked out systems,
there can be a difference).
* Add an ACB_RESET flag, so that a BUS DEVICE RESET can be queued within the
driver.
* If ACB_ABORT or ACB_RESET is set during reselection, schedule a message and
assert ATN.
* Optimize aic_datain_pio(), aic_dataout_pio() and aicintr() somewhat.
* Schedule a timeout only when we select the target, so that commands can't
time out prematurely.
influence from the old driver (written by Stefan Grefen) and from
an autochanger driver written by the Systems Programming Group at the
University of Utah Computer Science Department (currently residing
in src/sys/arch/hp300/dev/ac.c).
1) If we get an unexpected disconnect, issue a REQUEST SENSE, as recommended
by the SCSI-2 spec. If the target created a contingent allegiance condition,
this will clear it. Also, if it happened while sending a SDTR or WDTR message,
disable negotiation for that target.
2) Since some lame devices still don't deal correctly, make sure we deassert
ATN if our last message out is interrupted. If we get a MESSAGE PARITY ERROR,
we'll reassert ATN anyway. This should ensure that we never have to send a
MESSAGE NO OPERATION.
3) Set AIC_ABORTING only when actually sending a BUS DEVICE RESET or ABORT,
so we get better error detection.
4) Other internal reorganization of no consequence.
- Remove unnecessary prototypes
- Fix ioctl arguments
- Don't use kprintf(...TO_LOG...); directly. This does not set the log
priority, and uses priority of the previous message. Use log() with %:
(soon to be documented on mailing lists; eventually in section 9 manual
pages), most importantly:
(1) support interrupt pin swizzling on non-i386 systems with
PCI-PCI bridges (per PPB spec; done, but meaningless, on i386).
(2) provide pci_{io,mem}_find(), to determine what I/O or memory
space is described by a given PCI configuration space
mapping register.
(3) provide pci_intr_map(), pci_intr_string(), and
pci_intr_{,dis}establish() to manipulate and print info about
PCI interrupts.
(4) make pci functions take as an argument a machine-dependent
cookie, to allow more flexibility in implementation.
(1) use pci_{io,mem}_find(), to determine what I/O or memory
space is described by a given PCI configuration space
mapping register, and bus_{io,mem}_map() to map it.
(2) use pci_intr_map(), pci_intr_string(), and
pci_intr_{,dis}establish() to manipulate and print info about
PCI interrupts.
(5) make pci functions take as an argument a machine-dependent
cookie, to allow more flexibility in implementation.
(1) make pci functions take as an argument a machine-dependent
cookie, to allow more flexibility in implementation.
(use of other PCI functions, etc., left unchanged.)
to {mainbus,tc,ioasic}_cd.
Change ioasic config name from "lance " to "lance". Correct for
pmaxes, perhaps not for Alphas.
Boots and runs under load on a 5000/200. Hangs during boot on an ioasic
decstation; pmax autoconfig needs fixing.
Add Decstation-3100 baseboard support (untested).
* Display as much revision info as we can get.
* Fix a race condition that could cause interrupts to be lost.
* Handle `out of mailbox' conditions (much) more gracefully.
* Schedule timeouts more safely.
* Add diagnostic code to (correctly) detect if a CCB times out before its
outgoing mailbox is emptied.
the "Fast Assembler Version of the copy loop)
- Yet another sanity check included: initialize unused receive buffers to
zero sender id; and test for sender hardware id of zero in the receiver.
Gives us better error messages in case flaky hardware causes spurious
receive interupts.
change tc_submatch() to compile with the new device-attach scheme:
the TC bus uses a 'submatch' function which checks device locators
and then calls a match function.
Instead of calling cf->cf_driver->cd_match(), we now need to call
cf->cf_attach->ca_match().
Rename the extern declaration of `tccd' and `ioasiccd' in
src/sys/dev/tc/{tvar.h,ioasicvar.h", respectively, to be
`tc_cd' and `ioasic_cd', to match the 1.1B-style definitions.
only include the relevant code in the probe & attach functions. Still
one probe and one attach function, with #ifdefs, but this is a step
in the right direction and saves a few hundred bytes (ooh, ahh!).
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
- split softc size and match/attach out from cfdriver into
a new struct cfattach.
- new "attach" directive for files.*. May specify the name of
the cfattach structure, so that devices may be easily attached
to parents with different autoconfiguration semantics.
they're base I/O ports, not really addresses. rather than using those
softc fields directly, assign local variables (asicbase, nicbase) to have
their values, and use those. (consistency, plus better optimization
potential in some cases.)
eventually. rename the probe functions for the various boards to 'find' for
consistency/correctness (but the generic8390 check is still a 'probe').
Add protos for ed_find(), ed_probe_generic8390(), ed_find_WD80x3(),
ed_find_3Com(), and ed_find_Novell(). Fix one minor space nit.
'flags 1' on the sb? kernel configuration file line (because it frobs a
noncontiguous IO port to configure the Jazz16 extensions).
Also, remove static sb_device structure and fill in user's buffer on
each request.
ends of lines, turn aligned blocks of eight spaces into tabs, name all
softc elements with the "sc_" prefix, and call the softc by the commonly
used name, "sc", when passing it around.
pcibus and pci.
(2) remove the #ifdef i386 from pci.c, and provide a machine-dependent
hook (pci_md_attach_hook()) to do any machine-dependent attachment
gunk, e.g. on the i386 printing out the configuration mode (if bus 0)
(3) don't pass max device number for a given bus in, use
PCI_MAX_DEVICE_NUMBER, which can be defined on a per-machine basis.
(defaults to 32. on i386, it's 32 if pci conf mode == 1, 16 if 2.)
assumes that pci_map_mem/pci_map_io provide interfaces which utilitize
bus_{io,mem}_handle_t's, or types which are compatible. This works on the
i386, and will change eventually anyway.)
for U_INT8, INT16, U_INT16 definitions. Convert structs and
definitions in ncr_reg.h (e.g. ncrcmd, the chip register layout, etc.)
to use these definitions.
Add INB_OFF, INL_OFF, and OUTL_OFF macros to access specified offsets into
I/O or memory space. Convert register dumps (etc.), and cache snoop
test to use these new macros, so that nothing accesses the device
I/O or memory space directly. (Register dumps now come from I/O space
if NCR_IOMAPPED. They used to bogusly use memory space.)
Add a new relocation type for script entries, RELOC_KVAR. Allow scripts
access to mono_time.tv_sec, mono_time, and ncr_cache via this
mechanism, and convert scripts to use it. An ncrcmd is only 32 bits
wide, and KVAs may be > 32 bits wide (e.g. on Alpha), leading to
linker problems. This is a safer way to do this anyway; relocation is
more deterministic this way, and doesn't rely on KVAs not looking like
other relocation types.
Panic if an unmatched relocation other than 'zero' is specified. That's
now a script bug. (This used to be used to convert KVAs of kernel
variables referenced in the script to PCI bus physical addresses,
and that is now handled by RELOC_KVAR relocations.)
Figure out and print the model of chip.
assumes that pci_map_mem/pci_map_io provide interfaces which utilitize
bus_{io,mem}_handle_t's, or types which are compatible. This works on the
i386, and will change eventually anyway.)
substantial reworking of the multi-port drivers, as they need to frob
bits in the io-port spaces of their children. As a result, the
commulti->com attachment interface is substantially more complex.
(This may be fixable in the future by making some of the code common,
but as long as io-port allocation checking is planned, it's necessary.)
macro where appropriate. No point in hard-coding numbers in multiple places.
At the very least, this makes the drivers slightly easier to diff/understand.
attaching, and to the devices when attaching them. #include <machine/bus.h>
to make this backward compatible with old #include requirements.
Also, clean up idempotency so that isa/eisa/pci "var.h" headers are
consistent (make them all idempotent).
and will probably work with Massbus adapters as well. (Not tested,
but it's the same code as for 11/780). Ubareset's may cause crashes
on 8600 also like 11/780, but they are more uncommon. No support
for console RL02 yet, but it's likely to come.
but unconfigured (or unsupported) devices printed at boot time.
The device list is woefully incomplete, because i didn't put much time
in it, and because i don't know good "human names" for many TC devices.
and attach the secondary pci bus as a 'pci' device. Note that this support
is incomplete and will not yet work for ports other than that i386. (The
i386 can rely on the PCI interrupt 'line' information to determine
interrupt mapping, which is not necessarily possible on other systems.)
planned changes in other sections of code.
(2) expands the use of config_found() in tcattach() in-line, so that
tcsubmatch can be invoked there, rather than being invoked by
individual drivers.
(3) tcbus_attach_args now includes a turbochannel speed flag.
(4) compare a string in tcbus_attach_args against tc bus driver name,
to better allow attachment of multiple busses to the same parent.
keep going.
Improve error messages to include device name.
use STATIC for function declarations/definitions so it's easier to get a
useful symbol table by flipping a switch.
fix semantics of AUDIO_WERROR (now returns a count).
Also repair audio driver back to "real-time mode" where user must
provide data quickly enough for real time audio or silence is played
until user catches up. Add optional "play all" mode (additional bit in
mode field) to play all samples provided, with silence when needed to
avoid buffer underrun.
* Snap the sample rate when setting it, and remember only the time constant.
* Set the time constant when changing between play/record.
* Always return the actual sample rate with AUDIO_GETINFO.
* Add more delays while writing registers.
* Replace sc_dma{in,out}_inprogress with sc_dmadir.
* Eliminate the need for sc_locked.
* Add more DPRINTF()s.
detatch everywhere else. Reverted to the english spelling.
Also fixed the rest of the prototype warnings while I was at it.
ic/ncr5380sbc.c: Don't declare Debugger()... I have to clean this
everywhere :-(
Rename the ioctl asic register and slot macros from ASIC_<xxx> to
IOASIC_<xxx>, to be compatible with the machine-indpendent names in
sys/dev/tc/ioasicvar.h. The pmax code still uses
sys/arch/pmax/pmax/asic.h, as some of the registers and offsets
defined there are not yet defined in sys/dev/tc/ioasicvar.h.
Rename the ioctl asic base-address pointer from `asic_base' to `ioasic_base'.
Use the device address in the attach_args structure, instead of
using the deprecated BUS_CVTADDR macro.
Change the Mach derived asc driver to use "SCSI_PHASE_xxx" instead of
"ASC_PHASE_xxx", as the latest version of the Mach driver does.
cgd