DragonFlyBSD

On Wed, Feb 23, 2011 at 7:18 PM, Magliano Andre'
<masterblaster@tiscali.it> wrote:

Hi all,

i just discovered it's not possible with github to view a full diff
patch just clicking the commit
(http://support.github.com/discussions/api/227-no-diff-of-addedremoved-files-in-commit-show-api),
but with the suggested url:

https://github.com/{user}/{repo}/compare/{revision}^...{revision}.diff

so the patch gain as a bz2 download (far faster than my previous github
link):
https://github.com/downloads/randy1/DragonFlyBSD/ea7ae20415c0af1882c4.diff.bz2

I will take a look at it. Thanks for submission

Best Regards,
sephe

Hi Sepherosa,

thanks for your interest!

Just discovered yesterday, i was using the gpe init wrong; now i did it
as fbsd does, ie:

s/AcpiEnableGpe/AcpiSetGpeWakeMask/

and

@ -1627,6 +1628,9 @ acpi_probe_child(ACPI_HANDLE handle, UINT32 level,
void *context, void **status)
if (*search != NULL)
break;

+ if (type ACPI_TYPE_DEVICE && acpi_parse_prw(handle, &prw)
0)
+ AcpiSetupGpeForWake(handle, prw.gpe_handle, prw.gpe_bit);
+
/* * Create a placeholder device for this node. Sort the * placeholder so that the probe/attach passes will run

I'm seeing then in console output (filtered with 'grep Wake') with
ACPI set debug layer 'ACPI_EVENTS' level 'ACPI_LV_FUNCTIONS
ACPI_LV_INTERRUPTS'

evxfgpe-0375 [07] SetupGpeForWake       : ----Entry
 evxfgpe-0423 [07] SetupGpeForWake       : ----Exit- AE_OK
 evxfgpe-0375 [03] SetupGpeForWake       : ----Entry
 evxfgpe-0423 [03] SetupGpeForWake       : ----Exit- AE_OK
 evxfgpe-0375 [03] SetupGpeForWake       : ----Entry
 evxfgpe-0423 [03] SetupGpeForWake       : ----Exit- AE_OK
 evxfgpe-0375 [03] SetupGpeForWake       : ----Entry
 evxfgpe-0423 [03] SetupGpeForWake       : ----Exit- AE_OK
 evxfgpe-0375 [03] SetupGpeForWake       : ----Entry
 evxfgpe-0423 [03] SetupGpeForWake       : ----Exit- AE_OK
 evxfgpe-0375 [03] SetupGpeForWake       : ----Entry
 evxfgpe-0423 [03] SetupGpeForWake       : ----Exit- AE_OK
 evxfgpe-0375 [03] SetupGpeForWake       : ----Entry
 evxfgpe-0423 [03] SetupGpeForWake       : ----Exit- AE_OK
 evxfgpe-0457 [14] SetGpeWakeMask        : ----Entry
 evxfgpe-0508 [14] SetGpeWakeMask        : ----Exit- AE_OK
 evxfgpe-0457 [16] SetGpeWakeMask        : ----Entry
 evxfgpe-0508 [16] SetGpeWakeMask        : ----Exit- AE_OK

so it looks better now...

What it's definitively wrong:

*) acpi_tz0: _TMP value is absurd, ignored (2104.0C)

*) some kind of interrupt storm (snip below repeated very fast), slowing
down the whole system:

evsci-0154 [136] EvSciXruptHandler : ----Entry
evevent-0309 [136] EvFixedEventDetect : Fixed Event Block: Enable
00000120 Status 00000011
evgpe-0525 [136] EvGpeDetect : Read GPE Register at GPE08:
Status=00, Enable=02
evsci-0175 [136] EvSciXruptHandler : ----Exit- 0 0

gpe init from console log:

acpi_ec0: <Embedded Controller: GPE 0x9> [tentative] port 0x66,0x62 on
acpi0
evxface-0763 [05] InstallGpeHandler : ----Entry
evxface-0848 [05] InstallGpeHandler : ----Exit- AE_OK
evxfregn-0162 [05] InstallAddressSpaceHan: ----Entry
evregion-0974 [06] EvInstallSpaceHandler : ----Entry
evregion-0635 [08] EvDetachRegion : ----Entry
evregion-0652 [08] EvDetachRegion : ----Exit-
evregion-0777 [08] EvAttachRegion : ----Entry
evregion-0801 [08] EvAttachRegion : ----Exit- AE_OK
evregion-1178 [06] EvInstallSpaceHandler : ----Exit- AE_OK
evregion-1204 [06] EvExecuteRegMethods : ----Entry
evregion-0368 [08] EvExecuteRegMethod : ----Entry
evregion-0379 [08] EvExecuteRegMethod : ----Exit- AE_OK
evregion-1217 [06] EvExecuteRegMethods : ----Exit- AE_OK
evxfregn-0233 [05] InstallAddressSpaceHan: ----Exit- AE_OK
evxfgpe-0209 [05] EnableGpe : ----Entry
evgpe-0251 [06] EvAddGpeReference : ----Entry
evgpe-0156 [07] EvUpdateGpeEnableMask : ----Entry
evgpe-0178 [07] EvUpdateGpeEnableMask : ----Exit- AE_OK
evgpe-0201 [07] EvEnableGpe : ----Entry
evgpe-0227 [07] EvEnableGpe : ----Exit- AE_OK
evgpe-0276 [06] EvAddGpeReference : ----Exit- AE_OK
evxfgpe-0223 [05] EnableGpe : ----Exit- AE_OK
acpi_ec0: <Embedded Controller: GPE 0x9> [attached!] port 0x66,0x62 on
acpi0

I can't interpret this, as i have no knowledge of acpi at all (i'm
trying to read last acpi standard...). Could you maybe give me a hint?

Thanks!

On 2/24/2011, ""Sepherosa Ziehau \(via DragonFly issue tracker\)"
bugs@crater.dragonflybsd.org" <"Sepherosa Ziehau \(via DragonFly issue
tracker\)"
bugs@crater.dragonflybsd.org> wrote:

Sepherosa Ziehau <sepherosa@gmail.com> added the comment:

On Wed, Feb 23, 2011 at 7:18 PM, Magliano Andre'
<masterblaster@tiscali.it> wrote:

Hi all,

i just discovered it's not possible with github to view a full diff
patch just clicking the commit
(http://support.github.com/discussions/api/227-no-diff-of-addedremoved-files-in-commit-show-api),
but with the suggested url:

https://github.com/{user}/{repo}/compare/{revision}^...{revision}.diff

so the patch gain as a bz2 download (far faster than my previous github
link):
https://github.com/downloads/randy1/DragonFlyBSD/ea7ae20415c0af1882c4.diff.bz2

I will take a look at it. Thanks for submission

Best Regards,
sephe

----------
status: unread -> chatting

_____________________________________________
DragonFly issue tracker <bugs@lists.dragonflybsd.org>
<http://bugs.dragonflybsd.org/issue2003>
_____________________________________________

Hi Sepherosa,

got some additional information (hope it helps).

After this hunk (from fbsd) the overrall system reaction is ok (say, like no loading of acpi.ko)

diff --git a/sys/dev/acpica5/acpi_ec.c b/sys/dev/acpica5/acpi_ec.c
index 432cb5a..7d4f02e 100644
--- a/sys/dev/acpica5/acpi_ec.c
++ b/sys/dev/acpica5/acpi_ec.c
@ -658,6 +658,14 @ EcGpeQueryHandler(void Context)
device_printf(sc->ec_dev, "evaluation of query method %s failed: %s\n",
qxx, AcpiFormatException(Status));
}

+ / Reenable runtime GPE if its execution was deferred. */
+ if (sc->ec_sci_pend) {
+ Status = AcpiFinishGpe(sc->ec_gpehandle, sc->ec_gpebit);
+ if (ACPI_FAILURE(Status))
+ device_printf(sc->ec_dev, "reenabling runtime GPE failed: %s\n",
+ AcpiFormatException(Status));
+ }
}

/*

I tried booting a kernel with same debug level (events+function), but without the acpi code upgrade and i have an interrupt storm as well...

kernel bootlog with no acpi upgrade (i.e. acpi 20090521):

evsci-0154 [136] EvSciXruptHandler     : ----Entry
 evevent-0309 [136] EvFixedEventDetect    : Fixed Event Block: Enable 00000120 Status 00000011
   evgpe-0525 [136] EvGpeDetect           : Read GPE Register at GPE08: Status=00, Enable=02
   evsci-0175 [136] EvSciXruptHandler     : ----Exit-        0       0

kernel bootlog with acpi upgrade:

evsci-0154 [136] EvSciXruptHandler     : ----Entry
 evevent-0297 [137] EvFixedEventDetect    : ----Entry
 evevent-0309 [137] EvFixedEventDetect    : Fixed Event Block: Enable 00000120 Status 00000011
 evevent-0336 [137] EvFixedEventDetect    : ----Exit-        0       0
   evgpe-0525 [136] EvGpeDetect           : Read GPE Register at GPE00: Status=00, Enable=01
   evgpe-0525 [136] EvGpeDetect           : Read GPE Register at GPE08: Status=00, Enable=0E
   evsci-0175 [136] EvSciXruptHandler     : ----Exit-        0       0

the GPE registers may be interpreted in a different way in the new acpi release, but GPE00 register is now seen...

On Thu, 24 Feb 2011 02:39:55 +0000
"Sepherosa Ziehau \(via DragonFly issue tracker\)" <bugs@crater.dragonflybsd.org> wrote:

Sepherosa Ziehau <sepherosa@gmail.com> added the comment:

On Wed, Feb 23, 2011 at 7:18 PM, Magliano Andre'
<masterblaster@tiscali.it> wrote:

Hi all,

i just discovered it's not possible with github to view a full diff
patch just clicking the commit
(http://support.github.com/discussions/api/227-no-diff-of-addedremoved-files-in-commit-show-api),
but with the suggested url:

https://github.com/{user}/{repo}/compare/{revision}^...{revision}.diff

so the patch gain as a bz2 download (far faster than my previous github
link):
https://github.com/downloads/randy1/DragonFlyBSD/ea7ae20415c0af1882c4.diff.bz2

I will take a look at it. Thanks for submission

Best Regards,
sephe

----------
status: unread -> chatting

_____________________________________________
DragonFly issue tracker <bugs@lists.dragonflybsd.org>
<http://bugs.dragonflybsd.org/issue2003>
_____________________________________________

Hi Sepherosa,

please have a look (of course if you have time) at
https://github.com/randy1/DragonFlyBSD/commits/master/

It runs ok now, no system slow down anymore (and temp value correctly reported, finally!).

I'm anyway not sure if i'm using the GPE API correctly. I'll test it on another laptop, as the one i'm working on gets C3 disabled by default (PIIX chipset bug). I'm really looking forward...

Acpi userland utilities are coming soon.

On Thu, 24 Feb 2011 02:39:55 +0000
"Sepherosa Ziehau \(via DragonFly issue tracker\)" <bugs@crater.dragonflybsd.org> wrote:

Sepherosa Ziehau <sepherosa@gmail.com> added the comment:

On Wed, Feb 23, 2011 at 7:18 PM, Magliano Andre'
<masterblaster@tiscali.it> wrote:

Hi all,

i just discovered it's not possible with github to view a full diff
patch just clicking the commit
(http://support.github.com/discussions/api/227-no-diff-of-addedremoved-files-in-commit-show-api),
but with the suggested url:

https://github.com/{user}/{repo}/compare/{revision}^...{revision}.diff

so the patch gain as a bz2 download (far faster than my previous github
link):
https://github.com/downloads/randy1/DragonFlyBSD/ea7ae20415c0af1882c4.diff.bz2

I will take a look at it. Thanks for submission

Best Regards,
sephe

----------
status: unread -> chatting

_____________________________________________
DragonFly issue tracker <bugs@lists.dragonflybsd.org>
<http://bugs.dragonflybsd.org/issue2003>
_____________________________________________

I will take care of it after the mptable work