Re: swapcontext rewrite broke some software

[David Xu]

On 2013/04/16 21:24, Oliver Pinter wrote:

Hi!

After this commit:

commit ac0cfc7fcb1b51ee6aeacfd676fa6dfbe11eefb5
Author: davidxu 
Date:   Wed Apr 10 02:40:03 2013 +

 swapcontext wrapper can not be implemented in C, the stack pointer saved in
 the context becomes invalid when the function returns, same as setjmp,
 it must be implemented in assemble language, see discussions in PR
 misc/177624.

Some* software not found the swapcontext functions after this commit.
Please add a sentence to UPDATING file and/or bump the
__FreeBSD_version to reflect this change.


* qemu



Hi,

The change is reverted.

Regards,
David Xu

___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

IPv6 bind fails with 49 (#define EADDRNOTAVAIL 49 /* Can't assign requested address */)

[Sreenivasa Honnur]

I have a ipv6 interface(ping6 to a remove ipv6 works) when I try to bind to 
this address through a socket program "sobind" fails with "49" as return value. 
If I give "saddr6.sin6_addr =  in6addr_any;" sobind works.

Any idea what could be going wrong here? 

roundhay# ifconfig cxgbe1
cxgbe1: flags=8843 metric 0 mtu 1500

options=6c07bb
ether 00:07:43:11:89:88
inet6 2010::102 prefixlen 64
inet6 fe80::207:43ff:fe11:8988%cxgbe1 prefixlen 64 scopeid 0xd
nd6 options=21
media: Ethernet 10Gbase-SR 
status: active
roundhay# ping6 2010::101
PING6(56=40+8+8 bytes) 2010::102 --> 2010::101
16 bytes from 2010::101, icmp_seq=0 hlim=64 time=0.950 ms
16 bytes from 2010::101, icmp_seq=1 hlim=64 time=0.158 ms
^C
--- 2010::101 ping6 statistics ---
2 packets transmitted, 2 packets received, 0.0% packet loss
round-trip min/avg/max/std-dev = 0.158/0.554/0.950/0.396 ms



Code:
=
{
rv = socreate(AF_INET6, &sock, SOCK_STREAM, IPPROTO_TCP,
td->td_ucred, td);

if (rv != 0) {
os_log_error("sock create ipv6 %s failed %d.\n",
tbuf, rv);
return NULL;
}

family = saddr6.sin6_family = AF_INET6;
inet_pton(AF_INET6, "2010::102", &saddr6.sin6_addr);
saddr6.sin6_port = htons(ep->port);
saddr6.sin6_len = sizeof(struct sockaddr_in6);

rv = sobind(sock, (struct sockaddr *)&saddr6, td);
}


Here is the code snippet from where EADDRNOTAVAIL  is returned.

File : netinet6/in6_pcb.c

in6_pcbbind(register struct inpcb *inp, struct sockaddr *nam,
struct ucred *cred)
{
……..
……..

} else if (!IN6_IS_ADDR_UNSPECIFIED(&sin6->sin6_addr)) {
struct ifaddr *ifa;

sin6->sin6_port = 0;/* yech... */
if ((ifa = ifa_ifwithaddr((struct sockaddr *)sin6)) ==
NULL &&
(inp->inp_flags & INP_BINDANY) == 0) {
return (EADDRNOTAVAIL);  it fails here since 
ifa_ifwithaddr() returns NULL
}
……..
……..

}

File: net/if.c
struct ifaddr *
ifa_ifwithaddr(struct sockaddr *addr)
{

return (ifa_ifwithaddr_internal(addr, 1));
}

ifa_ifwithaddr_internal(struct sockaddr *addr, int getref)
{
……..
……..

TAILQ_FOREACH(ifp, &V_ifnet, if_link) {
IF_ADDR_RLOCK(ifp);
printf("ifp->xname:%s ifp->dname:%s\n",ifp->if_xname, 
ifp->if_dname);
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
if (ifa->ifa_addr->sa_family != addr->sa_family)   
ifa->ifa_addr->sa_family=18 & addr->sa_family=28. Even though the interface is 
in IPV6 mode its sa_family is not  set properly, 
continue;
}
Ifa = NULL;
……..
……..

}

___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: ipfilter(4) needs maintainer

[Ed Maste]

On 15 April 2013 16:12, Cy Schubert  wrote:
> The existing license isn't that BSD-friendly either, which is why it lives
> in contrib/. I think the 5.1.X GPLv2 is about the same friendliness as
> Darren's IPF 4.1.X license. As long as it's not in GENERIC should be fine.
> A person can always load it anyway.

There's a plan[1] to remove the remaining GPL components from base
over time.  Updating to the last ipfilter that's under the current
license is probably the path forward, unless it moves out to ports.

[1] https://wiki.freebsd.org/GPLinBase

-Ed
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: ipfilter(4) needs maintainer

[Cy Schubert]

In message 
, Ed Maste writes:
> On 15 April 2013 16:12, Cy Schubert  wrote:
> > The existing license isn't that BSD-friendly either, which is why it lives
> > in contrib/. I think the 5.1.X GPLv2 is about the same friendliness as
> > Darren's IPF 4.1.X license. As long as it's not in GENERIC should be fine.
> > A person can always load it anyway.
> 
> There's a plan[1] to remove the remaining GPL components from base
> over time.  Updating to the last ipfilter that's under the current
> license is probably the path forward, unless it moves out to ports.
> 
> [1] https://wiki.freebsd.org/GPLinBase

That's been pointed out to me. IPF's build/install scripts place header 
files in /usr/include, IMO unacceptable for a port. Going forward we go to 
4.1.34 (still under the old license) then look at options.



-- 
Cheers,
Cy Schubert 
FreeBSD UNIX: Web:  http://www.FreeBSD.org

The need of the many outweighs the greed of the few.


___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

Sigh.

Someone's broken the bus or interrupt handling code between -9 and
-head. Or ACPI, for all I know. It may even be something to do with
RFKILL.

Can you please boot verbosely, both on -9 and -head? you don't have to
do anything - just capture the logfile /var/run/dmesg.boot and attach
them here.

Hopefully something glaringly different will stand out.

Thanks,



adrian

On 18 April 2013 07:05, Artyom Mirgorodskiy
 wrote:
> No interrupts at all:
>
>
>
> interrupt total rate
>
> irq1: atkbd0 474 1
>
> irq9: acpi0 95 0
>
> irq22: ehci1 504 1
>
> irq23: ehci0 1037 3
>
> irq256: hpet0:t0 8426 30
>
> irq257: hpet0:t1 1823 6
>
> irq258: hpet0:t2 1992 7
>
> irq259: hpet0:t3 2112 7
>
> irq264: hdac0 118 0
>
> irq265: ahci0 4026 14
>
> Total 20607 75
>
>
>
>
>
> On Thursday 18 April 2013 06:56:14 Adrian Chadd wrote:
>
>> I've just tested -HEAD on a WB197 (AR9287 + bluetooth.) It works fine.
>
>>
>
>> I think it's very likely something to do with the bus enumeration. Can
>
>> you do a vmstat -i, see if you've seen any ath interrupts?
>
>>
>
>>
>
>>
>
>> adrian
>
> --
>
> Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

Ok, the relevant / interesting bit:s. John, any ideas?

HEAD:

pcib2:  irq 18 at device 28.2 on pci0
pcib0: allocated type 3 (0xe050-0xe05f) for rid 20 of pcib2
pcib2:   domain0
pcib2:   secondary bus 10
pcib2:   subordinate bus   10
pcib2:   memory decode 0xe050-0xe05f
pcib2:   no prefetched decode
pci10:  on pcib2
pci10: domain=0, physical bus=10
found-> vendor=0x168c, dev=0x002e, revid=0x01
domain=0, bus=10, slot=0, func=0
class=02-80-00, hdrtype=0x00, mfdev=0
cmdreg=0x0006, statreg=0x0010, cachelnsz=16 (dwords)
lattimer=0x00 (0 ns), mingnt=0x00 (0 ns), maxlat=0x00 (0 ns)
intpin=a, irq=10

^-- notice how this is irq 10

powerspec 3  supports D0 D1 D3  current D0
MSI supports 1 message
map[10]: type Memory, range 64, base 0xe050, size 16, enabled
pcib2: allocated memory range (0xe050-0xe050) for rid 10 of pci0:10:0:0
pcib2: matched entry for 10.0.INTA
pcib2: slot 0 INTA hardwired to IRQ 18
ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
ath0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps
ath0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps
24Mbps 36Mbps 48Mbps 54Mbps
ath0: 2T2R
ath0: 11ng MCS 20MHz
ath0: MCS 0-7: 6.5Mbps - 65Mbps
ath0: MCS 8-15: 13Mbps - 130Mbps
ath0: AR9287 mac 384.2 RF5133 phy 15.15
ath0: 2GHz radio: 0x; 5GHz radio: 0x00c0
ath0: Use hw queue 1 for WME_AC_BE traffic
ath0: Use hw queue 0 for WME_AC_BK traffic
ath0: Use hw queue 2 for WME_AC_VI traffic
ath0: Use hw queue 3 for WME_AC_VO traffic
ath0: Use hw queue 8 for CAB traffic
ath0: Use hw queue 9 for beacons
ath0: using multicast key search

versus

pcib2:  irq 18 at device 28.2 on pci0
pcib0: allocated type 3 (0xe050-0xe05f) for rid 20 of pcib2
pcib2:   domain0
pcib2:   secondary bus 10
pcib2:   subordinate bus   10
pcib2:   memory decode 0xe050-0xe05f
pcib2:   no prefetched decode
pci10:  on pcib2
pci10: domain=0, physical bus=10
found-> vendor=0x168c, dev=0x002e, revid=0x01
domain=0, bus=10, slot=0, func=0
class=02-80-00, hdrtype=0x00, mfdev=0
cmdreg=0x0006, statreg=0x0010, cachelnsz=16 (dwords)
lattimer=0x00 (0 ns), mingnt=0x00 (0 ns), maxlat=0x00 (0 ns)
intpin=a, irq=7

^-- notice this is irq 7

powerspec 3  supports D0 D1 D3  current D0
MSI supports 1 message
map[10]: type Memory, range 64, base 0xe050, size 16, enabled
pcib2: allocated memory range (0xe050-0xe050) for rid 10 of pci0:10:0:0
pcib2: matched entry for 10.0.INTA
pcib2: slot 0 INTA hardwired to IRQ 18
ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
ath0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps
ath0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps
24Mbps 36Mbps 48Mbps 54Mbps
ath0: 2T2R
ath0: 11ng MCS 20MHz
ath0: MCS 0-7: 6.5Mbps - 65Mbps
ath0: MCS 8-15: 13Mbps - 130Mbps
ath0: AR9287 mac 384.2 RF5133 phy 15.15
ath0: Use hw queue 1 for WME_AC_BE traffic
ath0: Use hw queue 0 for WME_AC_BK traffic
ath0: Use hw queue 2 for WME_AC_VI traffic
ath0: Use hw queue 3 for WME_AC_VO traffic
ath0: Use hw queue 8 for CAB traffic
ath0: Use hw queue 9 for beacons
ath0: using multicast key search
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[John Baldwin]

On Thursday, April 18, 2013 11:22:38 am Adrian Chadd wrote:
> Ok, the relevant / interesting bit:s. John, any ideas?

Only that this means absolutely nothing?  These are the values the BIOS wrote 
into the registers which we use as hints about whether or not ACPI lies about 
which interrupts are used when APIC is disabled.  The actually useful message
shows the same interrupts used in both cases:

> HEAD:
> 
> pcib2: slot 0 INTA hardwired to IRQ 18
> ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
> ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> 
> versus
> 
> pcib2: slot 0 INTA hardwired to IRQ 18
> ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
> ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60

-- 
John Baldwin
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

... Why would it differ for the same machine, different kernel?


Adrian


Sent from my Palm Pre on AT&T
On Apr 18, 2013 8:40 AM, John Baldwin  wrote: 

On Thursday, April 18, 2013 11:22:38 am Adrian Chadd wrote:

> Ok, the relevant / interesting bit:s. John, any ideas?



Only that this means absolutely nothing?  These are the values the BIOS wrote 

into the registers which we use as hints about whether or not ACPI lies about 

which interrupts are used when APIC is disabled.  The actually useful message

shows the same interrupts used in both cases:



> HEAD:

> 

> pcib2: slot 0 INTA hardwired to IRQ 18

> ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 
on pci10

> ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60

> 

> versus

> 

> pcib2: slot 0 INTA hardwired to IRQ 18

> ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 
on pci10

> ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60



-- 

John Baldwin


___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Artyom Mirgorodskiy]

I tried to check out revision 245031 (only sys/dev/ath) and got a working WiFi. 
I'll try to find a broken revision.

On Thursday 18 April 2013 11:37:17 John Baldwin wrote:
> On Thursday, April 18, 2013 11:22:38 am Adrian Chadd wrote:
> > Ok, the relevant / interesting bit:s. John, any ideas?
> 
> Only that this means absolutely nothing?  These are the values the BIOS wrote 
> into the registers which we use as hints about whether or not ACPI lies about 
> which interrupts are used when APIC is disabled.  The actually useful message
> shows the same interrupts used in both cases:
> 
> > HEAD:
> > 
> > pcib2: slot 0 INTA hardwired to IRQ 18
> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> > 
> > versus
> > 
> > pcib2: slot 0 INTA hardwired to IRQ 18
> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> 
> 
-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

On head?



Adrian



Sent from my Palm Pre on AT&T
On Apr 18, 2013 10:06 AM, Artyom Mirgorodskiy 
 wrote: 


I tried to check out revision 245031 (only sys/dev/ath) and got a working WiFi. 
I'll try to find a broken revision.
 
On Thursday 18 April 2013 11:37:17 John Baldwin wrote:
> On Thursday, April 18, 2013 11:22:38 am Adrian Chadd wrote:
> > Ok, the relevant / interesting bit:s. John, any ideas?
> 
> Only that this means absolutely nothing?  These are the values the BIOS 
wrote 
> into the registers which we use as hints about whether or not ACPI lies 
about 
> which interrupts are used when APIC is disabled.  The actually useful 
message
> shows the same interrupts used in both cases:
> 
> > HEAD:
> > 
> > pcib2: slot 0 INTA hardwired to IRQ 18
> > ath0:  mem 0xe050-0xe050 irq 18 at device 
0.0 on pci10
> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> > 
> > versus
> > 
> > pcib2: slot 0 INTA hardwired to IRQ 18
> > ath0:  mem 0xe050-0xe050 irq 18 at device 
0.0 on pci10
> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> 
> 
-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Artyom Mirgorodskiy]

Yes

On Thursday 18 April 2013 10:30:33 Adrian Chadd wrote:


On head?



Adrian




Sent from my Palm Pre on AT&T


Artyom MirgorodskiyOn Apr 18, 2013 10:06 AM,   
wrote: 



I tried to check out revision 245031 (only sys/dev/ath) and got a working WiFi. 
I'll try to find a broken revision.

On Thursday 18 April 2013 11:37:17 John Baldwin wrote:
> On Thursday, April 18, 2013 11:22:38 am Adrian Chadd wrote:
> > Ok, the relevant / interesting bit:s. John, any ideas?
> 
> Only that this means absolutely nothing? These are the values the BIOS wrote 
> into the registers which we use as hints about whether or not ACPI lies about 
> which interrupts are used when APIC is disabled. The actually useful message
> shows the same interrupts used in both cases:
> 
> > HEAD:
> > 
> > pcib2: slot 0 INTA hardwired to IRQ 18
> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> > 
> > versus
> > 
> > pcib2: slot 0 INTA hardwired to IRQ 18
> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on pci10
> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> 
> 
-- 
Artyom Mirgorodskiy


-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

Hm, so -HEAD everything else except the wifi code?

That's really odd. I updated to -HEAD this morning just to test the
AR9287 for you and it was peachy!


Adrian

On 18 April 2013 10:36, Artyom Mirgorodskiy
 wrote:
> Yes
>
>
>
> On Thursday 18 April 2013 10:30:33 Adrian Chadd wrote:
>
> On head?
>
>
>
> Adrian
>
>
>
> Sent from my Palm Pre on AT&T
>
> 
>
> Artyom MirgorodskiyOn Apr 18, 2013 10:06 AM, 
> wrote:
>
> I tried to check out revision 245031 (only sys/dev/ath) and got a working
> WiFi. I'll try to find a broken revision.
>
>
>
> On Thursday 18 April 2013 11:37:17 John Baldwin wrote:
>
>> On Thursday, April 18, 2013 11:22:38 am Adrian Chadd wrote:
>
>> > Ok, the relevant / interesting bit:s. John, any ideas?
>
>>
>
>> Only that this means absolutely nothing? These are the values the BIOS
>> wrote
>
>> into the registers which we use as hints about whether or not ACPI lies
>> about
>
>> which interrupts are used when APIC is disabled. The actually useful
>> message
>
>> shows the same interrupts used in both cases:
>
>>
>
>> > HEAD:
>
>> >
>
>> > pcib2: slot 0 INTA hardwired to IRQ 18
>
>> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on
>> > pci10
>
>> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
>
>> >
>
>> > versus
>
>> >
>
>> > pcib2: slot 0 INTA hardwired to IRQ 18
>
>> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on
>> > pci10
>
>> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
>
>>
>
>>
>
> --
>
> Artyom Mirgorodskiy
>
>
> --
>
> Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Artyom Mirgorodskiy]

Yes, HEAD everything else except the wifi code (ath).
Last working revision is 247286. So problem in revision 247287.


On Thursday 18 April 2013 10:42:15 Adrian Chadd wrote:
> Hm, so -HEAD everything else except the wifi code?
> 
> That's really odd. I updated to -HEAD this morning just to test the
> AR9287 for you and it was peachy!
> 
> 
> Adrian
> 
> On 18 April 2013 10:36, Artyom Mirgorodskiy
>  wrote:
> > Yes
> >
> >
> >
> > On Thursday 18 April 2013 10:30:33 Adrian Chadd wrote:
> >
> > On head?
> >
> >
> >
> > Adrian
> >
> >
> >
> > Sent from my Palm Pre on AT&T
> >
> > 
> >
> > Artyom MirgorodskiyOn Apr 18, 2013 10:06 AM, 
> > wrote:
> >
> > I tried to check out revision 245031 (only sys/dev/ath) and got a working
> > WiFi. I'll try to find a broken revision.
> >
> >
> >
> > On Thursday 18 April 2013 11:37:17 John Baldwin wrote:
> >
> >> On Thursday, April 18, 2013 11:22:38 am Adrian Chadd wrote:
> >
> >> > Ok, the relevant / interesting bit:s. John, any ideas?
> >
> >>
> >
> >> Only that this means absolutely nothing? These are the values the BIOS
> >> wrote
> >
> >> into the registers which we use as hints about whether or not ACPI lies
> >> about
> >
> >> which interrupts are used when APIC is disabled. The actually useful
> >> message
> >
> >> shows the same interrupts used in both cases:
> >
> >>
> >
> >> > HEAD:
> >
> >> >
> >
> >> > pcib2: slot 0 INTA hardwired to IRQ 18
> >
> >> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on
> >> > pci10
> >
> >> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> >
> >> >
> >
> >> > versus
> >
> >> >
> >
> >> > pcib2: slot 0 INTA hardwired to IRQ 18
> >
> >> > ath0:  mem 0xe050-0xe050 irq 18 at device 0.0 on
> >> > pci10
> >
> >> > ioapic0: routing intpin 18 (PCI IRQ 18) to lapic 0 vector 60
> >
> >>
> >
> >>
> >
> > --
> >
> > Artyom Mirgorodskiy
> >
> >
> > --
> >
> > Artyom Mirgorodskiy
-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[John Baldwin]

On Thursday, April 18, 2013 12:41:53 pm Adrian Chadd wrote:
> 
> ... Why would it differ for the same machine, different kernel?

I can't tell you why, but if you compare the full dmesg's you will see
that several devices all changed their BIOS-assigned IRQs because the BIOS
decided to shuffle the IRQs assigned to the PCI links around.  That should
all be moot since you are using the APIC anyway.

-- 
John Baldwin
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

On 18 April 2013 11:58, Artyom Mirgorodskiy
 wrote:
> Yes, HEAD everything else except the wifi code (ath).
>
> Last working revision is 247286. So problem in revision 247287.

That's really odd. It behaves perfectly fine on my system.

You've tested 247287 and it breaks?


Adrian
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Artyom Mirgorodskiy]

Yes. I tested 247287 and it breaks.

On Thursday 18 April 2013 13:31:29 Adrian Chadd wrote:
> On 18 April 2013 11:58, Artyom Mirgorodskiy
>  wrote:
> > Yes, HEAD everything else except the wifi code (ath).
> >
> > Last working revision is 247286. So problem in revision 247287.
> 
> That's really odd. It behaves perfectly fine on my system.
> 
> You've tested 247287 and it breaks?
> 
> 
> Adrian
-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

On 18 April 2013 13:33, Artyom Mirgorodskiy
 wrote:
> Yes. I tested 247287 and it breaks.

Hm.

Well, not much changed there.

Try going to 247287 (ie, the broken revision), but revert
head/sys/dev/ath/ath_hal/ar5416/ar5416_xmit.c to 247286.

See if that fixes it.



Adrian
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Artyom Mirgorodskiy]

Did not work.

On Thursday 18 April 2013 13:58:23 Adrian Chadd wrote:
> On 18 April 2013 13:33, Artyom Mirgorodskiy
>  wrote:
> > Yes. I tested 247287 and it breaks.
> 
> Hm.
> 
> Well, not much changed there.
> 
> Try going to 247287 (ie, the broken revision), but revert
> head/sys/dev/ath/ath_hal/ar5416/ar5416_xmit.c to 247286.
> 
> See if that fixes it.
> 
> 
> 
> Adrian
-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

On 18 April 2013 14:43, Artyom Mirgorodskiy
 wrote:
> Did not work.
>

Hm. Ok.

Let's add some debugging:

inside of ar5416SetChainMasks(), let's add a printf() at the -end- of
the function:

ath_hal_printf(ah, "%s: txchainmask=0x%x, rxchainmask=0x%x\n",
__func__, AH5416(ah)->ah_tx_chainmask, AH5416(ah)->ah_rx_chainmask);

Then recompile, reboot, and show me the output of 'dmesg'
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

On 18 April 2013 15:07, Artyom Mirgorodskiy
 wrote:
> See attached
>

What the hell? Those masks are really wrong.

Try adding this line after it:

ath_hal_printf(ah, "%s: pcap rx=0x%x, tx=0x%x; configured rx=0x%x,
tx=0x%x\n", __func__, pCap->halRxChainMask, pCap->halTxChainMask,
rx_chainmask, tx_chainmask);
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

Hm! I wonder..

Edit if_athvar.h - change sc_rxchainmask and sc_txchainmask in
ath_softc to be uint32_t, rather than int.

See if that helps.



Adrian

On 18 April 2013 15:20, Artyom Mirgorodskiy
 wrote:
> See attached
>
>
>
> On Thursday 18 April 2013 15:08:36 Adrian Chadd wrote:
>
>> On 18 April 2013 15:07, Artyom Mirgorodskiy
>
>>  wrote:
>
>> > See attached
>
>> >
>
>>
>
>> What the hell? Those masks are really wrong.
>
>>
>
>> Try adding this line after it:
>
>>
>
>> ath_hal_printf(ah, "%s: pcap rx=0x%x, tx=0x%x; configured rx=0x%x,
>
>> tx=0x%x\n", __func__, pCap->halRxChainMask, pCap->halTxChainMask,
>
>> rx_chainmask, tx_chainmask);
>
> --
>
> Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

On 18 April 2013 15:54, Artyom Mirgorodskiy
 wrote:
> Nothing :(

Ok, so I wonder now whether we're actually getting the right chainmask
at startup.

edit if_ath.c, look for 'ath_hal_getrxchainmask()'

After the rx/tx chainmask is fetched,a dd this:

device_printf(sc->sc_dev, "%s: RX chainmask=0x%x, TX
chainmask=0x%x\n", sc->sc_rxchainmask, sc->sc_txchainmask);



Adrian
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: Atheros 9287 - no carrier . revision 249623.

[Artyom Mirgorodskiy]

Nothing :(

On Thursday 18 April 2013 15:22:49 Adrian Chadd wrote:
> Hm! I wonder..
> 
> Edit if_athvar.h - change sc_rxchainmask and sc_txchainmask in
> ath_softc to be uint32_t, rather than int.
> 
> See if that helps.
> 
> 
> 
> Adrian
> 
> On 18 April 2013 15:20, Artyom Mirgorodskiy
>  wrote:
> > See attached
> >
> >
> >
> > On Thursday 18 April 2013 15:08:36 Adrian Chadd wrote:
> >
> >> On 18 April 2013 15:07, Artyom Mirgorodskiy
> >
> >>  wrote:
> >
> >> > See attached
> >
> >> >
> >
> >>
> >
> >> What the hell? Those masks are really wrong.
> >
> >>
> >
> >> Try adding this line after it:
> >
> >>
> >
> >> ath_hal_printf(ah, "%s: pcap rx=0x%x, tx=0x%x; configured rx=0x%x,
> >
> >> tx=0x%x\n", __func__, pCap->halRxChainMask, pCap->halTxChainMask,
> >
> >> rx_chainmask, tx_chainmask);
> >
> > --
> >
> > Artyom Mirgorodskiy
-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Cannot unmount nullfs in current

[Craig Rodrigues]

Hi,

I am trying to build some software which uses
nanobsd, and mounts/unmounts many nullfs mounts
while it runs.  I am hitting failures where
I cannot unmount nullfs file systems.  I cannot figure out why.

Here is more info.

SYSTEM
==
I am running amd64, current build at this revision:

10.0-CURRENT FreeBSD 10.0-CURRENT #0 r249181: Sat Apr  6 03:07:32 UTC 2013
amd64

STEPS TO REPRODUCE
===

(1)  Create a directory, /opt2/branches.  Make sure that /opt2/branches
 is on ZFS

(2)
   mkdir -p /opt2/branches/freenas
   mkdir -p /opt2/branches/freenas-cache

(3)

git clone git://github.com/freenas/freenas.git /opt2/branches/freenas
git clone git://github.com/freenas/ports.git/opt2/branches/freenas-cache/ports
git clone git://github.com/trueos/trueos.git/opt2/branches/freenas-cache/trueos

(4)  sudo to root

(5)  cd /opt2/branches/freenas

(6)
script build.log env GIT_REPO=/opt2/branches/freenas-cache/trueos \
GIT_PORTS_REPO=/opt2/branches/freenas-cache/ports \
sh build/do_build.sh


The build cranks for a while, and then I get this error:

00:02:37 ### log:
/opt2/branches/freenas/os-base/amd64/_.cust.add_pkg_archivers_lzo2
do_build.sh: ERROR: FreeNAS /opt2/branches/freenas/nanobsd/os-base build
FAILED; please check above log for more details



If I look in .cust.add_pkg_archivers_lzo2, I see this error:

+ umount /opt2/branches/freenas/os-base/amd64/_.w/usr/ports/distfiles
umount: unmount of
/opt2/branches/freenas/os-base/amd64/_.w/usr/ports/distfiles failed: Device
busy


If I try to do this manually:

# umount /opt2/branches/freenas/os-base/amd64/_.w/usr/ports/distfiles
umount: unmount of
/opt2/branches/freenas/os-base/amd64/_.w/usr/ports/distfiles failed: Device
busy


I can't figure out why this mount is busy.
If I do:

umount -f /opt2/branches/freenas/os-base/amd64/_.w/usr/ports/distfiles

it unmounts, but I don't like using the '-f' flag to force the unmount.

Any ideas?  I am attaching some of my logs.

--
Craig
Script started on Thu Apr 18 14:29:00 2013
command: env GIT_REPO=/opt2/branches/freenas-cache/trueos 
GIT_PORTS_REPO=/opt2/branches/freenas-cache/ports sh build/do_build.sh
Using local mirror in /opt2/branches/freenas-cache/trueos.
Using local git ports mirror in /opt2/branches/freenas-cache/ports
Sourcing /opt2/branches/freenas/nanobsd/os-base
Package is not yet built: ftp_wget
Package is not yet built: benchmarks_iozone
Package is not yet built: benchmarks_iperf
Package is not yet built: benchmarks_netperf
Package is not yet built: benchmarks_xdd
Package is not yet built: security_sudo
Package is not yet built: sysutils_ipmitool
Package is not yet built: www_py-django-json-rpc
Package is not yet built: devel_py-mimeparse
Package is not yet built: devel_py-six
Package is not yet built: devel_py-dateutil
Package is not yet built: devel_py-rose
Package is not yet built: www_py-django-tastypie
Package is not yet built: devel_py-daemon
Package is not yet built: devel_py-polib
Package is not yet built: devel_py-ujson
Package is not yet built: devel_py-simplejson
Package is not yet built: sysutils_bsdstats
Package is not yet built: www_wgetpaste
Package is not yet built: devel_py-greenlet
Package is not yet built: net_py-eventlet
Package is not yet built: graphics_jpeg
Package is not yet built: security_ca_root_nss
Package is not yet built: ftp_curl
Package is not yet built: devel_apr1
Package is not yet built: www_neon29
Package is not yet built: devel_subversion
Package is not yet built: editors_vim-lite
Package is not yet built: misc_py-pexpect
Package is not yet built: devel_ipython
Package is not yet built: devel_p5-Term-ReadKey
Package is not yet built: devel_p5-subversion
Package is not yet built: mail_p5-Net-SMTP-SSL
Package is not yet built: lang_p5-Error
Package is not yet built: devel_git
Package is not yet built: devel_ctags
Automatically building a * * F A T * * image so we can build ports
00:00:00 # NanoBSD image FreeNAS-9.1.0-ALPHA-4dd41d9_dirty-x64 build starting
00:00:00 ## Skipping buildworld (as instructed)
00:00:00 ## Skipping buildkernel (as instructed)
00:00:00 ## Clean and create world directory 
(/opt2/branches/freenas/os-base/amd64/_.w)
00:00:25 ## Construct install make.conf 
(/opt2/branches/freenas/os-base/amd64/make.conf.install)
00:00:25 ## installworld
00:00:25 ### log: /opt2/branches/freenas/os-base/amd64/_.iw
00:02:29 ## install /etc
00:02:29 ### log: /opt2/branches/freenas/os-base/amd64/_.etc
00:02:30 ## configure nanobsd /etc
00:02:30 ## install kernel (/opt2/branches/freenas/nanobsd/FREENAS.amd64)
00:02:30 ### log: /opt2/branches/freenas/os-base/amd64/_.ik
00:02:37 ## run customize scripts
00:02:37 ## [1/184] customize "clean_packages"
00:02:37 ### log: /opt2/branches/freenas/os-base/amd64/_.cust.clean_packages
00:02:37 ## Clean and create world directory 
(/opt2/branches/freenas/os-base/amd64/_.w/usr/local)
00:02:37 ## [2/184] customize "cust_install_files"
00:02:37 ### log: /opt2/branches/freenas/os-base/amd64/_.cust.c

Re: Atheros 9287 - no carrier . revision 249623.

[Adrian Chadd]

Ok. I'll add some tidyups to head tonight and then add some more verbose 
logging.

I don't have any 64 bit machines to test ar9287 on atm.

Sent from my Palm Pre on AT&T
On Apr 18, 2013 3:52 PM, Artyom Mirgorodskiy 
 wrote: 


Nothing :(
 
On Thursday 18 April 2013 15:22:49 Adrian Chadd wrote:
> Hm! I wonder..
> 
> Edit if_athvar.h - change sc_rxchainmask and sc_txchainmask in
> ath_softc to be uint32_t, rather than int.
> 
> See if that helps.
> 
> 
> 
> Adrian
> 
> On 18 April 2013 15:20, Artyom Mirgorodskiy
>  wrote:
> > See attached
> >
> >
> >
> > On Thursday 18 April 2013 15:08:36 Adrian Chadd wrote:
> >
> >> On 18 April 2013 15:07, Artyom Mirgorodskiy
> >
> >>  wrote:
> >
> >> > See attached
> >
> >> >
> >
> >>
> >
> >> What the hell? Those masks are really wrong.
> >
> >>
> >
> >> Try adding this line after it:
> >
> >>
> >
> >> ath_hal_printf(ah, "%s: pcap rx=0x%x, tx=0x%x; configured rx=0x%x,
> >
> >> tx=0x%x\n", __func__, pCap->halRxChainMask, 
pCap->halTxChainMask,
> >
> >> rx_chainmask, tx_chainmask);
> >
> > --
> >
> > Artyom Mirgorodskiy
-- 
Artyom Mirgorodskiy
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"