s/jumo/jumbo/ in comments.

Back out the following changes:

http://mail-index.netbsd.org/source-changes/2020/02/29/msg114581.html
http://mail-index.netbsd.org/source-changes/2020/02/29/msg114583.html

They have reportedly caused some issues with link detection, and will
be revisited later.

branches: 1.30.8; 1.30.12;
No functional change:
- s/u_int64_t/uint64_t/
- s/alignes/aligns/
- indent, tabify

s/impliment/implement/ in comment.

u_int*_t -> unit*_t. No funcitonal change.

miscelaneous -> miscellaneous in comment.

No functional change:

- Fix typo in comment.
- Whitespace

G/C the never-used version of ti_hostaddr that can't possibly be correct.

- Change the definition and usage of TI_HOSTADDR() to allow for 64-bit
DMA addresses. Modeled after similar usage in the bge(4) driver (the
chips supported by bge(4) are descendants of those supported by ti(4)).
- Use the 64-bit PCI DMA tag if available, otherwise we're allocating (and
using) bounce buffers needlessly.

Adopt <net/if_stats.h>.

branches: 1.22.2;
if_flags is neither int nor short. It's unsigned short.

branches: 1.21.4;
Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

branches: 1.20.30; 1.20.38;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory

Back out the following changes:

http://mail-index.netbsd.org/source-changes/2020/02/29/msg114581.html
http://mail-index.netbsd.org/source-changes/2020/02/29/msg114583.html

They have reportedly caused some issues with link detection, and will
be revisited later.

No functional change:
- s/u_int64_t/uint64_t/
- s/alignes/aligns/
- indent, tabify

s/impliment/implement/ in comment.

u_int*_t -> unit*_t. No funcitonal change.

miscelaneous -> miscellaneous in comment.

No functional change:

- Fix typo in comment.
- Whitespace

G/C the never-used version of ti_hostaddr that can't possibly be correct.

- Change the definition and usage of TI_HOSTADDR() to allow for 64-bit
DMA addresses. Modeled after similar usage in the bge(4) driver (the
chips supported by bge(4) are descendants of those supported by ti(4)).
- Use the 64-bit PCI DMA tag if available, otherwise we're allocating (and
using) bounce buffers needlessly.

Adopt <net/if_stats.h>.

branches: 1.22.2;
if_flags is neither int nor short. It's unsigned short.

branches: 1.21.4;
Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

branches: 1.20.30; 1.20.38;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory

No functional change:
- s/u_int64_t/uint64_t/
- s/alignes/aligns/
- indent, tabify

s/impliment/implement/ in comment.

u_int*_t -> unit*_t. No funcitonal change.

miscelaneous -> miscellaneous in comment.

No functional change:

- Fix typo in comment.
- Whitespace

G/C the never-used version of ti_hostaddr that can't possibly be correct.

- Change the definition and usage of TI_HOSTADDR() to allow for 64-bit
DMA addresses. Modeled after similar usage in the bge(4) driver (the
chips supported by bge(4) are descendants of those supported by ti(4)).
- Use the 64-bit PCI DMA tag if available, otherwise we're allocating (and
using) bounce buffers needlessly.

Adopt <net/if_stats.h>.

branches: 1.22.2;
if_flags is neither int nor short. It's unsigned short.

Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

branches: 1.20.38;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory

u_int*_t -> unit*_t. No funcitonal change.

miscelaneous -> miscellaneous in comment.

No functional change:

- Fix typo in comment.
- Whitespace

G/C the never-used version of ti_hostaddr that can't possibly be correct.

- Change the definition and usage of TI_HOSTADDR() to allow for 64-bit
DMA addresses. Modeled after similar usage in the bge(4) driver (the
chips supported by bge(4) are descendants of those supported by ti(4)).
- Use the 64-bit PCI DMA tag if available, otherwise we're allocating (and
using) bounce buffers needlessly.

Adopt <net/if_stats.h>.

branches: 1.22.2;
if_flags is neither int nor short. It's unsigned short.

Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

branches: 1.20.38;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory

G/C the never-used version of ti_hostaddr that can't possibly be correct.

- Change the definition and usage of TI_HOSTADDR() to allow for 64-bit
DMA addresses. Modeled after similar usage in the bge(4) driver (the
chips supported by bge(4) are descendants of those supported by ti(4)).
- Use the 64-bit PCI DMA tag if available, otherwise we're allocating (and
using) bounce buffers needlessly.

Adopt <net/if_stats.h>.

branches: 1.22.2;
if_flags is neither int nor short. It's unsigned short.

Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

branches: 1.20.38;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory

Adopt <net/if_stats.h>.

if_flags is neither int nor short. It's unsigned short.

Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

branches: 1.20.38;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory

if_flags is neither int nor short. It's unsigned short.

Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

branches: 1.20.38;
split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory

Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes
if_upgt.c's ETHER_ALIGN from 0 to 2.

split device_t/softc for all remaining drivers.
replace "struct device *" with "device_t".
use device_xname(), device_unit(), etc.

branches: 1.19.58; 1.19.68;
First part of ti(4) overhaul:
- Sanitise macros
- Apply KNF
Should have no functional changes.

branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.17.26;
merge ktrace-lwp.

branches: 1.16.2;
- const poisoning
- avoid variable shadowing.

nuke trailing whitespace

branches: 1.14.4; 1.14.6;
Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable).
Bump the default values for these to the values used by FreeBSD,
and also adjust ti_init_rx_ring_jumbo() to use the same constant
that FreeBSD uses. Yes, this consumes more kernel memory.

The effect of this is that you can use jumbo frames in a back-to-back
setup with TCP windows up to about 250KB and get ~930Mbit/s throughput,
while we were earlier limited to around 3-400Mbit/s, and trying to push
above that mark by widening the TCP window caused
ti0: jumbo buffer allocation failed
messages to be logged and a corresponding stall in the traffic.

branches: 1.13.2;
Use the bus_space_*_stream_* methods to access the chip's registers.
As we turn the chip to big-endian mode on big-endian systems, we should
never byte-swap the data read/written from/to registers. Tested on sparc64.
Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting
bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)

branches: 1.12.2;
coaless -> coalesce.

branches: 1.11.16;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

Add macros to sync the various bits of the DMA'd control data.

Use macros (like those in every other network driver I have written)
to compute the offset of a structure in the DMA'd control data, and
the DMA address of that structure.

Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions,
for the same reason we split the tx encap routine.

Split the transmit encap routine into two versions: one for the Tigon 1
and another for the Tigon 2. This eliminates a conditional in the
loop around the DMA segments, and will make further cleanup easier.

Move the `ti_tx_ring_nic' pointer out of the ring-data structure
and into the softc. Garbage-collect the `ti_cmd_ring' pointer,
since bus_space is properly used to access the command ring.

Replace a bunch of home-grown Ethernet related constants with
ones defined in <net/if_ether.h>

delint: () protect some macro args

branches: 1.3.2;
Pull up FreeBSD changes:
1.39:
Have if_ti stop "hiding" the softc pointer in the buffer region. Rather,
use the available void * passed to the free routine and pass the softc
pointer through there.
1.33:
Add support for the Netgear GA620T copper gigabit card.
1.32:
Tweak probe message so that 1000baseSX and 1000baseT cards are
explicitly identified.
1.31:
Update the Tigon driver to support 1000baseTX gigE over copper AceNIC
cards. This basically involves switching to the 12.4.13 firmware, plus
a couple of minor tweaks to the driver.

NetBSD changes:
get rid of ti_inuse, the mbuf ref counting code should call ti_free() when
needed.
Use hardware 802.1q support.

NBPG -> PAGE_SIZE

branches: 1.1.2; 1.1.4; 1.1.14;
quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver.
todo:
-IPv6
-clean up jumbo buffer allocation - NetBSD provides an opaque argument
to the free function, thus doesn't need the hack done here
-deal correctly with the mapping of the shared memory