CVS log for src/sys/dev/ic/dwc_gmac.c

Up to [cvs.NetBSD.org] / src / sys / dev / ic

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

spelling and grammar fixes, mainly in comments.

awge(4): Narrow scope of `core' lock and rename to mcast lock.

Sprinkle locking notes.

Proposed at:

https://mail-index.netbsd.org/source-changes-d/2024/08/10/msg014251.html

Comment-only changes from the updated patch at:

https://mail-index.netbsd.org/source-changes-d/2024/08/10/msg014252.html

awge(4): MP improvements

Remove the non-MP-safe scaffolding and make the locking less
coarse.

Sprinkle some const

Handle GMAC_MAC_Version's that include USERVER.

Improve and add new debug output

Fix some bus_dmamap_sync calls.
- ensure new descriptors are written before handing ownership of the
  first in the list to the device.
- ensure descriptors are sync'ed before dumping them in
  dwc_gmac_dump_[rt]x_desc
- don't sync the TX packet buffer twice (and not after we've passed
  ownership of its TX descriptor)

Change some variable names to better reflect what they are while I'm here.

sys: Drop redundant NULL check before m_freem(9)

m_freem(9) safely has accepted NULL argument at least since 4.2BSD:
https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c

Compile-tested on amd64/ALL.

Suggested by knakahara@

Misc whitespace

Use ether_crc32_be() instead of having a local bitrev32() function to
munge ether_crc32_le() output when programming multicast filter.

Reorder the bus_dmamap_sync sync operations

 BUS_DMASYNC_PREWRITE | BUS_DMASYNC_PREREAD to
 BUS_DMASYNC_PREREAD | BUS_DMASYNC_PREWRITE

for consistency.

More KNF (whitespace around binary operators)

Enforce the device 32 bit DMA limitation via bus_dmatag_subregion if bus
can address more than 4GB.

KNF - spaces around binary operators.

Fix spello in debug output

Remove unnecssary #include

fix simple mis-matched function prototype and definitions.

most of these are like, eg

   void foo(int[2]);

with either of these

   void foo(int*) { ... }
   void foo(int[]) { ... }

in some cases (such as stat or utimes* calls found in our header files),
we now match standard definition from opengroup.

found by GCC 12.

Eliminate use of IFF_OACTIVE.

Pull up following revision(s) (requested by sekiya in ticket #1755):

	sys/dev/ic/dwc_gmac.c: revision 1.76-1.77

Turn off AWIN_GMAC_MAC_CONF_ACS, so that all received packets retain FCS
bytes.
ether_input() can now trust M_HASFCS to accurately represent the packet
contents.

Discussed on tech-net@

Pull up following revision(s) (requested by sekiya in ticket #1499):

	sys/dev/ic/dwc_gmac.c: revision 1.76-1.77

Turn off AWIN_GMAC_MAC_CONF_ACS, so that all received packets retain FCS
bytes.
ether_input() can now trust M_HASFCS to accurately represent the packet
contents.

Discussed on tech-net@

Turn off AWIN_GMAC_MAC_CONF_ACS, so that all received packets retain FCS
bytes.

ether_input() can now trust M_HASFCS to accurately represent the packet
contents.

Discussed on tech-net@

Do not unilaterally set M_HASFCS; this breaks protocols that lack CRC bytes
(such as AppleTalk).

Instead, remove/preserve the final four bytes in the packet ourselves on a per-
protocol basis, as we do in arch/arm/xscale/ixp425_if_npe.c (and dev/ic/gem.c,
and so forth).

Add missing double p and d for stopped and overriden accordingly.
Fix few more typos along the way, mainly in copy-pasted comments.

Sync w/ HEAD.

if_attach and if_initialize cannot fail, don't test return value

These were originally made failable back in 2017 when if_initialize
allocated a softint in every interface for link state changes, so
that it could fail gracefully instead of panicking:

https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html

However, this spawned many seldom- or never-tested error branches,
which are risky to have around.  And that softint in every interface
has since been replaced by a single global workqueue, because link
state changes require thread context but not low latency or high
throughput:

https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html

So there is no longer any reason for if_initialize to fail.  (The
subroutine if_stats_init can't fail because percpu_alloc can't fail
either.)

There is a snag: the softint_establish in if_percpuq_create could
fail, potentially leading to bad consequences later on trying to use
the softint.  This change doesn't introduce any new bugs because of
the snag -- if_percpuq_attach was already broken.  However, the snag
can be better addressed without spawning error branches, either by
using a single softint or making softints less scarce.

(Separate commit will change the signatures of if_attach and
if_initialize to return void, scheduled to ride whatever is the next
convenient kernel bump.)

Patch and testing on amd64 and evbmips64-eb by maya@; commit message
soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.

sync with head

 Cast to uint32_t to avoid undefined behavior in dwc_gmac_write_hwaddr().
Found by kUBSan.

Sync w/ HEAD.

extend the timeout value. dwc_gmac_reset() sometimes takes more time.

Pull up following revision(s) (requested by mrg in ticket #1045):

	sys/kern/uipc_mbuf.c: revision 1.235
	sys/dev/ic/dwc_gmac.c: revision 1.70
	sys/dev/ic/dwc_gmac_reg.h: revision 1.20
	sys/dev/ic/dwc_gmac.c: revision 1.66
	sys/dev/ic/dwc_gmac.c: revision 1.67
	sys/dev/ic/dwc_gmac.c: revision 1.68

awge: fix issue that caused rx packets to be corrupt with DIAGNOSTIC kernel

It seems the hardware can only reliably do rx DMA to addresses that are
dcache size aligned. This is hinted at by some GMAC data sheets but hard to
find an authoritative source.

on non-DIAGNOSTIC kernels we always implicitly get MCLBYTES-aligned mbuf
data pointers, but with the reintroduction of POOL_REDZONE for DIAGNOSTIC
we can get 8-byte alignment due to redzone padding. So align rx pointers to
64 bytes which should be good for both arm32 and aarch64.
While here change some bus_dmamap_load() to bus_dmamap_load_mbuf() and add
one missing bus_dmamap_sync(). Also fixes the code to not assume that
MCLBYTES == AWGE_MAX_PACKET. User may override MCLSHIFT in kernel config.
correct pointer arithmetics

mcl_cache: align items to COHERENCY_UNIT

Because we do cache incoherent DMA to/from mbufs we cannot safely share
share cache lines with adjacent items that may be concurrently accessed.

awge: drop redundant m_adj(). Handled via uipc_mbuf.c r1.235 instead.

Mask all the MMC counter interrupts if the MMC module is present.

prop_data_data_nocopy -> prop_data_value

Mask all the MMC counter interrupts if the MMC module is present.

Mostly merge changes from HEAD upto 20200411

Merge changes from current as of 20200406

Sync with head.

Adopt <net/if_stats.h>.

awge: drop redundant m_adj(). Handled via uipc_mbuf.c r1.235 instead.

correct pointer arithmetics

awge: fix issue that caused rx packets to be corrupt with DIAGNOSTIC kernel

It seems the hardware can only reliably do rx DMA to addresses that are
dcache size aligned. This is hinted at by some GMAC data sheets but hard to
find an authoritative source.

on non-DIAGNOSTIC kernels we always implicitly get MCLBYTES-aligned mbuf
data pointers, but with the reintroduction of POOL_REDZONE for DIAGNOSTIC
we can get 8-byte alignment due to redzone padding. So align rx pointers to
64 bytes which should be good for both arm32 and aarch64.

While here change some bus_dmamap_load() to bus_dmamap_load_mbuf() and add
one missing bus_dmamap_sync(). Also fixes the code to not assume that
MCLBYTES == AWGE_MAX_PACKET. User may override MCLSHIFT in kernel config.

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

move DWCGMAC_MPSAFE into dwc_gmac_var.h and introduce an
additional define that is 0 or FDT_INTR_MPSAFE that the
frontends can use when passing to fdtbus_intr_establish().

with NET_MPSAFE enabled, this avoids hangs seen on rock64,
as well as finishing the MPSAFE port for this driver.

XXX: still don't know why the existing hangs occur.

 Add rnd(9) support.

Sync with HEAD

 Whitespace fix (mainly tabify).

-No functional change:
  - KNF
  - u_int*_t -> uint*_t.

Store IFF_ALLMULTI in ec_flags instead of if_flags to avoid data races

IFF_ALLMULTI is set/unset to if_flags via if_mcast_op.  To avoid data races on
if_flags, IFNET_LOCK was added for if_mcast_op.  Unfortunately it produces
a deadlock so we want to remove added IFNET_LOCK by avoiding the data races by
another approach.

This fix introduces ec_flags to struct ethercom and stores IFF_ALLMULTI to it.
ec_flags is protected by ETHER_LOCK and thus IFNET_LOCK is no longer necessary
for if_mcast_op.  Note that the fix is applied only to MP-safe drivers that
the data races matter.

In the kernel, IFF_ALLMULTI is set by a driver and used by the driver itself.
So changing the storing place doesn't break anything.  One exception is
ioctl(SIOCGIFFLAGS); we have to include IFF_ALLMULTI in a result if needed to
export the flag as well as before.

A upcoming commit will remove IFNET_LOCK.

PR kern/54189

Protect ether_multi list

Fix this driver because it can set IFEF_MPSAFE.

 Fix previous. Pass phy_id correctly.

Pass the phy ID to dwc_gmac_attach.

Sync with HEAD

 Change MII PHY read/write API from:

	int (*mii_readreg_t)(device_t, int, int);
	void (*mii_writereg_t)(device_t, int, int, int);
to:

	int (*mii_readreg_t)(device_t, int, int, uint16_t *);
	int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

 In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

 Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

	arm/at91/at91emac.c
	arm/ep93xx/epe.c
	arm/omap/omapl1x_emac.c
	mips/ralink/ralink_eth.c
	arch/powerpc/booke/dev/pq3etsec.c(read)
	dev/cadence/if_cemac.c <- hkenken
	dev/ic/lan9118.c


Tested with the following device:

	axe+ukphy
	axe+rgephy
	axen+rgephy	(tested by Andrius V)
	wm+atphy
	wm+ukphy
	wm+igphy
	wm+ihphy
	wm+makphy
	sk+makphy
	sk+brgphy
	sk+gentbi
	msk+makphy
	sip+icsphy
	sip+ukphy
	re+rgephy
	bge+brgphy
	bnx+brgphy
	gsip+gphyter
	rtk+rlphy
	fxp+inphy	(tested by Andrius V)
	tlp+acphy
	ex+exphy
	epic+qsphy
	vge+ciphy	(tested by Andrius V)
	vr+ukphy	(tested by Andrius V)
	vte+ukphy	(tested by Andrius V)

Not tested (MAC):
	arm:at91emac
	arm:cemac
	arm:epe
	arm:geminigmac
	arm:enet
	arm:cpsw
	arm:emac(omac)
	arm:emac(sunxi)
	arm:npe
	evbppc:temac
	macppc:bm
	macppc:gm
	mips:aumac
	mips:ae
	mips:cnmac
	mips:reth
	mips:sbmac
	playstation2:smap
	powerpc:tsec
	powerpc:emac(ibm4xx)
	sgimips:mec
	sparc:be
	sf
	ne(ax88190, dl10019)
	awge
	ep
	gem
	hme
	smsh
	mtd
	sm
	age
	alc
	ale
	bce
	cas
	et
	jme
	lii
	nfe
	pcn
	ste
	stge
	tl
	xi
	aue
	mue
	smsc
	udav
	url

Not tested (PHY):
	amhphy
	bmtphy
	dmphy
	etphy
	glxtphy
	ikphy
	iophy
	lxtphy
	nsphyter
	pnaphy
	rdcphy
	sqphy
	tlphy
	tqphy
	urlphy

Sync with head

Bring back support for enhanced descriptor format in newer core versions.
Fix a few endian bugs and check the main core version before trying to
read the hardware feature mask. Only read the hardware feature register
if core version reported is >= 3.5 (all my older hardware reports 0).
With some hints and patches from jared, and ok from aymeric

Ssync with HEAD

Revert the recent support for enhanced descriptors until a viable solution is
found.

It broke at least Cubietruck, Orange Pi One, and maybe Rock64 ethernets in
different ways. Unfortunately it works on my Olinuxino Micro and Lime 2 so I
can't reproduce the bugs for now.

Should fix PR#53637

Add support for the enhanced descriptors feature.

This makes "recent" dwc gmac controllers, as found e.g. on the Cyclone V, work.
The change was also tested working on an Allwinner A20 which doesn't have the
feature.

No negative reaction on port-arm.

Sync with HEAD

Be consistent among ethernet drivers on the convention for printing ethernet
addresses.

NFC

dwc_gmac_attach: return non-zero on failure

 Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward
compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same
as FreeBSD.

 This change also fixes a bug that the direction is misunderstand on some
environment by passing the direction to bpf_mtap*() instead of checking
m->m_pkthdr.rcvif.

Sync with HEAD

Write MAC address high register before low register. Apparently the
hardware updates the filter when the low register is written.

Clear IFF_RUNNING | IFF_OACTIVE when stopping interface.

Pull up following revision(s) (requested by jmcneill in ticket #885):

	sys/dev/ic/dwc_gmac.c: revision 1.47

Avoid calling bus_dmamap_sync with len=0

Avoid calling bus_dmamap_sync with len=0

Add flag for disabling store-and-forward mode, and a callback for notifying
bus glue that the link speed has changed.

Pull up following revision(s) (requested by martin in ticket #456):
	sys/dev/ic/dwc_gmac.c: 1.45
Don't KASSERT that we are MPSAFE if ! DWCGMAC_MPSAFE

Pull up following revision(s) (requested by ozaki-r in ticket #456):
	sys/arch/arm/sunxi/sunxi_emac.c: 1.9
	sys/dev/ic/dwc_gmac.c: 1.43-1.44
	sys/dev/pci/if_iwm.c: 1.75
	sys/dev/pci/if_wm.c: 1.543
	sys/dev/pci/ixgbe/ixgbe.c: 1.112
	sys/dev/pci/ixgbe/ixv.c: 1.74
	sys/kern/sys_socket.c: 1.75
	sys/net/agr/if_agr.c: 1.43
	sys/net/bpf.c: 1.219
	sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416
	sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257
	sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146
	sys/net/if_etherip.c: 1.40
	sys/net/if_ethersubr.c: 1.243, 1.246
	sys/net/if_faith.c: 1.57
	sys/net/if_gif.c: 1.132
	sys/net/if_l2tp.c: 1.15, 1.17
	sys/net/if_loop.c: 1.98-1.101
	sys/net/if_media.c: 1.35
	sys/net/if_pppoe.c: 1.131-1.132
	sys/net/if_spppsubr.c: 1.176-1.177
	sys/net/if_tun.c: 1.142
	sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121
	sys/net/npf/npf_ifaddr.c: 1.3
	sys/net/npf/npf_os.c: 1.8-1.9
	sys/net/rtsock.c: 1.230
	sys/netcan/if_canloop.c: 1.3-1.5
	sys/netinet/if_arp.c: 1.255
	sys/netinet/igmp.c: 1.65
	sys/netinet/in.c: 1.210-1.211
	sys/netinet/in_pcb.c: 1.180
	sys/netinet/ip_carp.c: 1.92, 1.94
	sys/netinet/ip_flow.c: 1.81
	sys/netinet/ip_input.c: 1.362
	sys/netinet/ip_mroute.c: 1.147
	sys/netinet/ip_output.c: 1.283, 1.285, 1.287
	sys/netinet6/frag6.c: 1.61
	sys/netinet6/in6.c: 1.251, 1.255
	sys/netinet6/in6_pcb.c: 1.162
	sys/netinet6/ip6_flow.c: 1.35
	sys/netinet6/ip6_input.c: 1.183
	sys/netinet6/ip6_output.c: 1.196
	sys/netinet6/mld6.c: 1.90
	sys/netinet6/nd6.c: 1.239-1.240
	sys/netinet6/nd6_nbr.c: 1.139
	sys/netinet6/nd6_rtr.c: 1.136
	sys/netipsec/ipsec_output.c: 1.65
	sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10
kmem_intr_free kmem_intr_[z]alloced memory
the underlying pools are the same but api-wise those should match
Unify IFEF_*_MPSAFE into IFEF_MPSAFE
There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.
Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).
Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.
Proposed on tech-kern@ and tech-net@
Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch
It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..."
scattered all over the source code and makes it easy to identify remaining
KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE.
No functional change
Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE
If IFEF_MPSAFE is set, hold the lock and otherwise don't hold.
This change requires additions of KERNEL_LOCK to subsequence functions from
if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe
components.
Proposed on tech-kern@ and tech-net@
Ensure to hold if_ioctl_lock when calling if_flags_set
Fix locking against myself on ifpromisc
vlan_unconfig_locked could be called with holding if_ioctl_lock.
Ensure to not turn on IFF_RUNNING of an interface until its initialization completes
And ensure to turn off it before destruction as per IFF_RUNNING's description
"resource allocated". (The description is a bit doubtful though, I believe the
change is still proper.)
Ensure to hold if_ioctl_lock on if_up and if_down
One exception for if_down is if_detach; in the case the lock isn't needed
because it's guaranteed that no other one can access ifp at that point.
Make if_link_queue MP-safe if IFEF_MPSAFE
if_link_queue is a queue to store events of link state changes, which is
used to pass events from (typically) an interrupt handler to
if_link_state_change softint. The queue was protected by KERNEL_LOCK so far,
but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt
handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it
by a spin mutex.
Additionally with this change KERNEL_LOCK of if_link_state_change softint is
omitted if NET_MPSAFE is enabled.
Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of
if_timer (see the comment).
Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH
At that point no other one modifies the list so IFADDR_READER_FOREACH
is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though,
if we try to detect contract violations of pserialize, using it violates
the contract. So avoid using it makes life easy.
Ensure to call if_addr_init with holding if_ioctl_lock
Get rid of outdated comments
Fix build of kernels without ether
By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that
created a unnecessary dependency from if.c to if_ethersubr.c.
PR kern/52790
Rename IFNET_LOCK to IFNET_GLOBAL_LOCK
IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then).
Wrap if_ioctl_lock with IFNET_* macros (NFC)
Also if_ioctl_lock perhaps needs to be renamed to something because it's now
not just for ioctl...
Reorder some destruction routines in if_detach
- Destroy if_ioctl_lock at the end of the if_detach because it's used in various
  destruction routines
- Move psref_target_destroy after pr_purgeif because we want to use psref in
  pr_purgeif (otherwise destruction procedures can be tricky)
Ensure to call if_mcast_op with holding IFNET_LOCK
Note that CARP doesn't deal with IFNET_LOCK yet.
Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held
Describe which lock is used to protect each member variable of struct ifnet
Requested by skrll@
Write a guideline for converting an interface to IFEF_MPSAFE
Requested by skrll@
Note that IFNET_LOCK must not be held in softint
Don't set IFEF_MPSAFE unless NET_MPSAFE at this point
Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.
Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.

Don't KASSERT that we are MPSAFE if ! DWCGMAC_MPSAFE

Don't set IFEF_MPSAFE unless NET_MPSAFE at this point

Because recent investigations show that interfaces with IFEF_MPSAFE need to
follow additional restrictions to work with the flag safely. We should enable it
on an interface by default only if the interface surely satisfies the
restrictions, which are described in if.h.

Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because
the network stack is still serialized by the big kernel locks by default.

Pull up following revision(s) (requested by msaitoh in ticket #427):
	sys/arch/amiga/dev/if_bah_zbus.c: 1.17
	sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30
	sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32
	sys/arch/usermode/dev/if_veth.c: 1.9
	sys/dev/ic/an.c: 1.66
	sys/dev/ic/athn.c: 1.17
	sys/dev/ic/atw.c: 1.162
	sys/dev/ic/bwi.c: 1.33
	sys/dev/ic/dwc_gmac.c: 1.41-1.42
	sys/dev/ic/malo.c: 1.10
	sys/dev/ic/rt2560.c: 1.31
	sys/dev/ic/rt2661.c: 1.36
	sys/dev/ic/rt2860.c: 1.29
	sys/dev/ic/rtw.c: 1.127
	sys/dev/ic/rtwvar.h: 1.46
	sys/dev/ic/smc90cx6.c: 1.71
	sys/dev/ic/smc90cx6var.h: 1.12
	sys/dev/ic/wi.c: 1.244
	sys/dev/pci/if_ipw.c: 1.66
	sys/dev/pci/if_iwi.c: 1.104
	sys/dev/pci/if_iwm.c: 1.76
	sys/dev/pci/if_iwn.c: 1.86
	sys/dev/pci/if_rtwn.c: 1.13
	sys/dev/pci/if_wm.c: 1.541
	sys/dev/pci/if_wpi.c: 1.79
	sys/dev/pci/ixgbe/ixgbe.c: 1.106
	sys/dev/pci/ixgbe/ixv.c: 1.73 via patch
	sys/dev/pcmcia/if_malo_pcmcia.c: 1.15
	sys/dev/scsipi/if_se.c: 1.95
	sys/dev/usb/if_upl.c: 1.60
	sys/net/if.c: 1.396
	sys/net/if.h: 1.241
	sys/net/if_arc.h: 1.23
	sys/net/if_arcsubr.c: 1.78
	sys/net/if_bridge.c: 1.136-1.137
	sys/net/if_etherip.c: 1.39
	sys/net/if_faith.c: 1.56
	sys/net/if_gif.c: 1.131
	sys/net/if_loop.c: 1.96
	sys/net/if_mpls.c: 1.30
	sys/net/if_pppoe.c: 1.129
	sys/net/if_srt.c: 1.27
	sys/net/if_stf.c: 1.102
	sys/net/if_tap.c: 1.100
	sys/net/if_vlan.c: 1.105
	sys/netinet/ip_carp.c: 1.91
	sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74
	sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56
 if_initalize() and if_attach() failed when resource allocation failed
(e.g. allocating softint). Without this change, it panics. It's bad because
resource shortage really occured when a lot of pseudo interface is created.
To avoid this problem, don't panic and change return value of if_initialize()
and if_attach() to int. Caller fanction will be recover from error cleanly by
checking the return value.
 Return if bah_attach_subr() failed.
 If if_attach() failed in the attach function, return.
- If if_initialize() failed in the attach function, free resources and return.
- Add some missing frees in bridge_clone_destroy().
- KNF
 If error occured in bcmeth_ccb_attach(), free resources and return.
 If error occured in pq3etsec_attach(), free resources and return.
 If error occured in the attach function, free resources and return.
- If if_initialize() failed in athn_attach(), free resources and return.
- Add missing pmf_event_deregister() in athn_detach().
- Free resources correctly on some errors in atw_attach().
- Use apint*() insread of printf() in the attach function.
 If if_initialize() failed in the attach function, return.
- If if_initialize() failed in the attach function, free resources and return.
- Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach
   failed.
- If if_initialize() failed in the attach function, free resources and return.
- ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc.
- If if_initialize() failed in the attach function, free resources and return.
- Fix error path in the attach function correctly.
 If if_initialize() failed in the attach function, free resources and return.
 If if_attach() failed in the attach function, free resources and return.
 - If if_initialize() failed in the attach function, free resources and return.
 - KNF
 - If if_attach() failed in the attach function, free resources and return.
 - KNF
 Fix compile error.
 Fix compile error.
We don't need '&mii', but just 'mii' for mii_detach().
Don't free sc_rthash twice

update from HEAD

file dwc_gmac.c was added on branch tls-maxphys on 2017-12-03 11:37:03 +0000

Unify IFEF_*_MPSAFE into IFEF_MPSAFE

There are already two flags for if_output and if_start, however, it seems such
MPSAFE flags are eventually needed for all if_XXX operations. Having discrete
flags for each operation is wasteful of if_extflags bits. So let's unify
the flags into one: IFEF_MPSAFE.

Fortunately IFEF_*_MPSAFE flags have never been included in any releases, so
we can change them without breaking backward compatibility of the releases
(though the kernel version of -current should be bumped).

Note that if an interface have both MP-safe and non-MP-safe operations at a
time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe
opeartions take the kernel lock.

Proposed on tech-kern@ and tech-net@

We don't need '&mii', but just 'mii' for mii_detach().

- If if_initialize() failed in the attach function, free resources and return.
- Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach
   failed.

Sync with HEAD

Sync with HEAD

Sync with HEAD

Apply deferred if_start to more drivers

Sync with HEAD

Use our ipq... sorry martin

Merge from nick-nhusb - some MPification

Sync with HEAD.  (Note that most of these changes are simply $NetBSD$
tag issues.)

Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input

The benefits of the change are:
- We can reduce codes
- We can provide the same behavior between drivers
  - Where/When if_ipackets is counted up
  - Note that some drivers still update packet statistics in their own
    way (periodical update)
- Moved bpf_mtap run in softint
  - This makes it easy to MP-ify bpf

Proposed on tech-kern and tech-net

Protect sc_if_flags in dwc_gmac_ifflags_cb with the lock

Define/assert IFEF_START_MPSAFE

Sync with HEAD

Fix indent botch that crept in

Oops... use the ipq we created.

Fix locking botch

First pass at making this driver and attachments NET_MPSAFE aware.

I've not tested this yet.

Introduce m_set_rcvif and m_reset_rcvif

The API is used to set (or reset) a received interface of a mbuf.
They are counterpart of m_get_rcvif, which will come in another
commit, hide internal of rcvif operation, and reduce the diff of
the upcoming change.

No functional change.

Sync with HEAD

Introduce softint-based if_input

This change intends to run the whole network stack in softint context
(or normal LWP), not hardware interrupt context. Note that the work is
still incomplete by this change; to that end, we also have to softint-ify
if_link_state_change (and bpf) which can still run in hardware interrupt.

This change softint-ifies at ifp->if_input that is called from
each device driver (and ieee80211_input) to ensure Layer 2 runs
in softint (e.g., ether_input and bridge_input). To this end,
we provide a framework (called percpuq) that utlizes softint(9)
and percpu ifqueues. With this patch, rxintr of most drivers just
queues received packets and schedules a softint, and the softint
dequeues packets and does rest packet processing.

To minimize changes to each driver, percpuq is allocated in struct
ifnet for now and that is initialized by default (in if_attach).
We probably have to move percpuq to softc of each driver, but it's
future work. At this point, only wm(4) has percpuq in its softc
as a reference implementation.

Additional information including performance numbers can be found
in the thread at tech-kern@ and tech-net@:
http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html

Acknowledgment: riastradh@ greatly helped this work.
Thank you very much!

Sync with HEAD

Pull up following revision(s) (requested by jmcneill in ticket #958):
	sys/dev/ic/dwc_gmac.c: revision 1.34
No need to BUS_DMASYNC_PREREAD the transmit data buffer. Fixes "NFS writes
being corrupted?" issue on ODROID-C1 as seen on port-arm.

No need to BUS_DMASYNC_PREREAD the transmit data buffer. Fixes "NFS writes
being corrupted?" issue on ODROID-C1 as seen on port-arm.

Supports ETHERCAP_VLAN_MTU. ok martin@ PR#49967

Sync with HEAD

Pull up following revision(s) (requested by martin in ticket #546):
	sys/dev/ic/dwc_gmac.c: revision 1.32
Apply patch from FUKAUMI Naoki to fix ring buffer handling when the
ring fills completely.

Apply patch from FUKAUMI Naoki to fix ring buffer handling when the
ring fills completely.

Pull up following revision(s) (requested by snj in ticket #481):
	sys/dev/ic/dwc_gmac_var.h: revision 1.6
	sys/dev/ic/dwc_gmac_reg.h: revision 1.13
	sys/dev/ic/dwc_gmac.c: revision 1.25
	sys/dev/ic/dwc_gmac.c: revision 1.26
	sys/dev/ic/dwc_gmac.c: revision 1.28
	sys/dev/ic/dwc_gmac.c: revision 1.31
- Add MII flow control support
- Set "disable jabber" and "auto padding/CRC stripping" bits in MAC conf
- Write intr mask to intr mask register, not status
- Setup both TX and RX burst modes
- Setup both TX and RX store & forward modes
- Correct an issue with the "end" descriptor passed to dwc_gmac_txdesc_sync
  in dwc_gmac_tx_intr
ok martin@
Revert previous change to dwc_gmac_txintr() - while it looks strange at
first sight (and sorry I didn't spot it when reviewing), it is a small
optimization and actually correct.
Add a comment explaining it.
After handling an interrupt, try to handle more packets as we may have
space in the descriptor ring now.
Pointed out by Jared.
fix a couple txq fencepost issues, from FUKAUMI Naoki <fun@naobsd.org>

fix a couple txq fencepost issues, from FUKAUMI Naoki <fun@naobsd.org>

Pull up following revision(s) (requested by martin in ticket #383):
	sys/dev/ic/dwc_gmac.c: revision 1.30
Fix index vs. count confusion when syncing a range of newly queued TX
descriptors. Pointed out by FUKAUMI Naoki in private chat.

Fix index vs. count confusion when syncing a range of newly queued TX
descriptors. Pointed out by "naobsd" in private chat.

Pull up following revision(s) (requested by jmcneill in ticket #300):
	sys/dev/ic/dwc_gmac.c: revision 1.29
if we dont have a mac address, make one up

if we dont have a mac address, make one up

After handling an interrupt, try to handle more packets as we may have
space in the descriptor ring now.
Pointed out by Jared.

Pull up following revision(s) (requested by jmcneill in ticket #259):
	sys/dev/ic/dwc_gmac.c: revision 1.27
Fix bus_dmamap_sync usage.

Fix bus_dmamap_sync usage.

Revert previous change to dwc_gmac_txintr() - while it looks strange at
first sight (and sorry I didn't spot it when reviewing), it is a small
optimization and actually correct.
Add a comment explaining it.

- Add MII flow control support
- Set "disable jabber" and "auto padding/CRC stripping" bits in MAC conf
- Write intr mask to intr mask register, not status
- Setup both TX and RX burst modes
- Setup both TX and RX store & forward modes
- Correct an issue with the "end" descriptor passed to dwc_gmac_txdesc_sync
  in dwc_gmac_tx_intr

ok martin@

Pull up following revision(s) (requested by martin in ticket #189):
	sys/dev/ic/dwc_gmac.c: revision 1.1-1.24
	sys/dev/ic/dwc_gmac_reg.h: revision 1.1-1.12
	sys/dev/ic/dwc_gmac_var.h: revision 1.1-1.5
Add support for Synopsis Designware GMAC ethernet core, as found on
various Allwiner boards and used by the awge(4) driver.

file dwc_gmac.c was added on branch netbsd-7 on 2014-11-09 19:06:57 +0000

Typo in comment

Typo

Greatly simplify ioctl handling by fully relying on ether_ioctl and a
ifflags-change callback, together with a shadow copy of the last active
interface flags.
Fix ALLMULTI vs. PROMISC handling.
Fix mac hash filter calculation by reversing the crc bits (pointed out
by Jared).

Fix typo. Use format string.

multicast hash filter support

Set M_HASFCS does the frame does include the FCS in the length.

add $NetBSD$ to top of files

Cosmetics - block interrupts while enabling different interrupt sources.

Consistently set the RX descriptors controll word - do not disable RX
interrupt and set chain mode. Now it does not run off the end of the
ring after 512 packets any more.

Cleanup debug code a bit.

if we read an insane mac address from the chip, dont attach

only set PR bit in frame filter if IFF_PROMISC is set, and set PM bit if IFF_ALLMULTI is set

fix inverted logic with AWIN_GMAC_MAC_CONF_MIISEL bit; my cubieboard2 can talk to the network now!

Add more MAC filter setup, some DMA burst configuration (from jmcneill),
actually enable RX interrupts (spotted by jmcneill), add RX handling
code and debug code.

Do not flush/restart the TX engine after queuing packets, but instead
keep it running (it will stall when out of descriptors), and inform it
about new descriptors available by writing to the TXPOLL request register.
Add more debugging code.

When link status changes, update MAC configuration accordingly.

More interrupt handling, more debug code, less magic numbers.

Add (disabled) debug code.
Fix DMA descriptor link initialization.

Cleanup and __BIT()ify

Simplify device property handling by moving it into the core driver.
Overriding the MAC address now works again.

Remove debug printf, now that MII status changes work

Make the MII clock variable and passed in from the frontend.

Fix MII setup and interrupt handling, from Robert Swindell.

Add work-in-progress driver for the Designware GMAC core, found on some
allwinner chips.