Re: interrupts on AMD hardware

[John J. Karcher]

On 05/04/2018 04:47 AM, Alexander Boettcher wrote:

Hi John,

On 03.05.2018 17:47, John J. Karcher wrote:

On 05/02/2018 07:42 AM, Alexander Boettcher wrote:

Ben & John,

may you please try my branch [0] with the run/ahci_bench and run/lwip
scenarios and tell me whether this change anything for your setups ?

Thanks,

Alex.

[0] https://github.com/alex-ab/genode/commits/acpica_ext

Thanks Alex for all that work!

Unfortunately, I am not advanced enough in Git or the new Genode build
tools to know how to do this.  Could you provide the commands needed to
download/merge and build?  (I have created a clean Sculpt EA tree, as
described in the tutorial.)

never mind, I got it now tested myself and actually could fix it
differently for Genode as VM in Virtualbox [0]. The drivers
(sata/network) of Genode will run in Virtualbox in the upcoming Sculpt TC.

Cheers,

Alex.

[0] https://github.com/genodelabs/genode/issues/2801


Great news - can't wait to try it!


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: interrupts on AMD hardware

[John J. Karcher]

On 05/02/2018 07:42 AM, Alexander Boettcher wrote:


Ben & John,

may you please try my branch [0] with the run/ahci_bench and run/lwip
scenarios and tell me whether this change anything for your setups ?

Thanks,

Alex.

[0] https://github.com/alex-ab/genode/commits/acpica_ext


Thanks Alex for all that work!

Unfortunately, I am not advanced enough in Git or the new Genode build 
tools to know how to do this.  Could you provide the commands needed to 
download/merge and build?  (I have created a clean Sculpt EA tree, as 
described in the tutorial.)


Thanks again!

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: interrupts on AMD hardware

[John J. Karcher]

On 04/12/2018 05:02 AM, Christian Helmuth wrote:

Ben,

On Thu, Apr 12, 2018 at 02:14:04AM -0600, Nobody III wrote:

What exactly are you referring to, and how do I do it?


Unfortunately, you need to dig through the code, but if you dare it
may be really rewarding. I'm neither the author of the Genode
acpica-library port nor an expert in ACPICA at a whole, but find and
grep yielded me some hints where to look at.

So, navigate to repos/libports/src/lib/acpica and try to find some
appropriate backend functions to instrument. This may not suffice, so
you should also navigate to contrib/acpica-/src/lib/acpica and
instrument source/components/executer/exregion.c
(AcpiExSystemMemorySpaceHandler). To be honest, from this point it's
up to you to invest more effort to get your platform running.

Good luck


Nobody, I am also interesting in looking at this code, with the goal of 
getting networking to work in a VirtualBox client.  (My main machine is 
AMD-based, so I am interested in that also, but I like to do most of my 
experimenting in a VM.)


Unfortunately, I don't have much free time at the moment, but I can try 
to assist your efforts as much as I can.  I haven't worked on 
driver-level code in years, but I don't mind shaking off the rust for a 
good cause.  ;^)


If you post your discoveries / observations on this ML (or elsewhere, as 
appropriate), I will do the same.


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: usb issues on AMD 9xx chipset

[John J. Karcher]

This is interesting.  I get the same error message in a VirtualBox VM 
using the "ICH9" chipset.  (It doesn't happen with the "PIIX3" chipset, 
but the block device isn't found using that.)


How did you disable "usb_drv"?  (I tried a few things, but they all hung 
the system.)



On 03/31/2018 02:44 PM, Nobody III wrote:
No. I can use platform_drv just fine without usb_drv, and even with 
fb_drv.


On Sat, Mar 31, 2018, 10:32 AM Chris Rothrock > wrote:


if you don't load the usb_drv, does the platform_drv still run out
of caps?

On Sat, Mar 31, 2018 at 2:46 AM, Nobody III
> wrote:

I've been testing Genode's current compatibility with my
desktop computer, which has an AMD 9xx chipset. While the
framebuffer driver works fine, usb_drv does not.

Here's the full output:

Genode 18.02-2-g6888094f6 
16299 MiB RAM and 63253 caps assigned to init
[init -> acpi_drv] Found MADT
[init -> acpi_drv] MADT IRQ 0 -> GSI 2 flags: 0
[init -> acpi_drv] MADT IRQ 9 -> GSI 9 flags: 15
[init -> acpi_drv] Found MCFG
[init -> acpi_drv] MCFG BASE 0xe000 seg 0x0 bus 0x0-0xff
[init -> platform_drv] 0:2.0 - bridge 1:0.0
[init -> platform_drv] 0:9.0 - bridge 2:0.0
[init -> platform_drv] 0:14.4 - bridge 3:0.0
[init -> platform_drv] 0:15.0 - bridge 4:0.0
[init -> platform_drv] 0:15.3 - bridge 5:0.0
[init -> usb_drv] Could not read screen resolution in config node
[init -> usb_drv] Configured HID screen with 0x0
(multitouch=false)
[init -> usb_drv] No  config node found - not
starting the USB Storage (Block) service
[init -> usb_drv] No  config node found - not starting
the USB Nic (Network) service
[init -> usb_drv] No  config node found - not starting
external USB service
[init -> usb_drv] Enabled UHCI (USB 1.0/1.1) support
[init -> usb_drv] Enabled OHCI (USB 1.0/1.1) support
[init -> usb_drv] Enabled EHCI (USB 2.0) support
[init -> usb_drv] Enabled XHCI (USB 3.0) support
*[init -> platform_drv] Warning: Out_of_caps during
Device_component construction*
*[init -> platform_drv] Warning: Out_of_caps during
Device_component construction*
*[ 0] sys_assign_pci: Invalid Hint (0x200)*
[init -> platform_drv] Error: usb_drv -> : assignment of PCI
device 2:0.0 failed phys=0xe020 virt=0x8040
[init -> usb_drv] ioremap: mapped phys 0xfe10 (size 32768)
to virt 0x2
[init -> usb_drv] dev_info: xHCI Host Controller
[init -> usb_drv] dev_info: new USB bus registered, assigned
bus number 1
[init -> usb_drv] dev_info: hcc params 0x040040a5 hci version
0x100 quirks 0x0010
[init -> platform_drv] 2:0.0 adjust IRQ as reported by ACPI:
10 -> 48
[init -> platform_drv] 2:0.0 uses MSI 64bit, vector 0x9f,
address 0xfee1, maskable
[init -> usb_drv] dev_info: USB hub found
[init -> usb_drv] dev_info: 4 ports detected
[init -> usb_drv] dev_info: xHCI Host Controller
[init -> usb_drv] dev_info: new USB bus registered, assigned
bus number 2
[init -> usb_drv] dev_info: We don't know the algorithms for
LPM for this host, disabling LPM.
[init -> usb_drv] dev_info: USB hub found
[init -> usb_drv] dev_info: 4 ports detected
[ 0] sys_assign_pci: Invalid Hint (0xb2)
[init -> platform_drv] Error: usb_drv -> : assignment of PCI
device 0:16.2 failed phys=0xe00b2000 virt=0x1000
[init -> usb_drv] ioremap: mapped phys 0xfe204000 (size 256)
to virt 0x17000
[init -> usb_drv] dev_info: EHCI Host Controller
[init -> usb_drv] dev_info: new USB bus registered, assigned
bus number 3
[init -> usb_drv] dev_info: applying AMD
SB700/SB800/Hudson-2/3 EHCI dummy qh workaround
[init -> platform_drv] 0:16.2 adjust IRQ as reported by ACPI:
11 -> 17
[init -> platform_drv] 0:16.2 uses IRQ, vector 0x11, non-maskable
[init -> usb_drv] dev_info: irq 130, io mem 0xfe204000
[init -> usb_drv] dev_info: USB 2.0 started, EHCI 1.00
[init -> usb_drv] dev_info: USB hub found
[init -> usb_drv] dev_info: 4 ports detected
[ 0] sys_assign_pci: Invalid Hint (0x9a)
[init -> platform_drv] Error: usb_drv -> : assignment of PCI
device 0:13.2 failed phys=0xe009a000 virt=0x1000
[init -> usb_drv] ioremap: mapped phys 0xfe207000 (size 256)
to virt 0x28000
[init -> usb_drv] dev_info: EHCI Host Controller
[init -> usb_drv] dev_info: new USB bus registered, assigned
bus number 4
[init -> usb_drv] dev_info: applying AMD

Re: Sculpt EA On VirtualBox

[John J. Karcher]

On 03/05/2018 02:48 PM, Alexander Boettcher wrote:

On 05.03.2018 17:07, John J. Karcher wrote:

On 03/05/2018 08:05 AM, Alexander Boettcher wrote:


Since I primarily run everything I do in VirtualBox (and don't have
the recommended hardware anyway), I created a VM as close to the
recommendations as possible.

It will work until the point where you try to start a VM in Sculpt,
which will not run because VirtualBox don't support nested
virtualization.

That's all I need!  Could you send me the configuration for a working
VM?

Sorry, you got me wrong, I haven't a working configuration. I just noted
that even if you get Sculpt running in a VM, it will be limited in use.


Sorry about that.

Is there anything I can try to get this running (either networking with 
ICH9, or storage with PIIX3)?


I don't mind trying hard-coded values in drivers, or something like that.

--
  John J. Karcher
devu...@alternateapproach.com

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: Sculpt EA On VirtualBox

[John J. Karcher]

On 03/05/2018 08:05 AM, Alexander Boettcher wrote:

Hi,

On 04.03.2018 08:22, John J. Karcher wrote:

Actually, while the problems at SourceForge were delaying the mailing
list, I solved part of my problem.  Details below...


On 03/01/2018 11:14 PM, John J. Karcher wrote:

Hello,

I have downloaded, built, and begun testing Sculpt EA, following the
instructions linked to in the 18.02 release notes.  Everything went
well during the build process.

Since I primarily run everything I do in VirtualBox (and don't have
the recommended hardware anyway), I created a VM as close to the
recommendations as possible.


It will work until the point where you try to start a VM in Sculpt,
which will not run because VirtualBox don't support nested virtualization.


That's all I need!  Could you send me the configuration for a working 
VM?  My virtual SATA drive is not recognized if using PIIX3. 
("VBoxManage showvminfo" is verbose, but we don't have to worry about 
missing anytihng.)



[drivers -> platform_drv] Error: runtimne -> nic -> nic: assignment of
PCI device 0:11.0 failed phys=0xdc088000 virt=0x1000


This is a normal message if you have no IOMMU support, which virtualbox
doesn't provide to guests. You may just ignore (and we have to silence
the output in the future.)


That's good to know.


...
[runtimne -> nic -> nic] bus_addr = e242 len = 2
...
[drivers -> platform_drv] 0:11.0 uses IRQ, vector 0xb, non-maskable



This is closer to the hardware than I usually get these days.  Is there
anything I can try?


It sounds like the issue reported in
https://github.com/genodelabs/genode/issues/1044 - unfortunately there
has been no solution for ICH9 and network, but a quirk for PCIIX3 and
network exists. Maybe you can report there your findings, if you have
some news.


Interesting.  The log output of "download.config" doesn't contain the 
network-related errors when using PIIX3 - hopefully that's a good sign. 
(Without storage, I can't tell if it actually works, of course.)


Thanks!

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: Sculpt EA On VirtualBox

[John J. Karcher]

Actually, while the problems at SourceForge were delaying the mailing 
list, I solved part of my problem.  Details below...



On 03/01/2018 11:14 PM, John J. Karcher wrote:

Hello,

I have downloaded, built, and begun testing Sculpt EA, following the 
instructions linked to in the 18.02 release notes.  Everything went well 
during the build process.


Since I primarily run everything I do in VirtualBox (and don't have the 
recommended hardware anyway), I created a VM as close to the 
recommendations as possible.  It booted fine from a virtual CD, the 
system came up using the VESA framebuffer driver, the mouse worked, and 
I was able to follow the instructions until I got to the part about 
storage.  Unfortunately, there are no "" nodes under 
"" in "/report/drivers/block_devices".


I have the VM set up with a "SATA" controller and a single virtual 
drive.  Has anyone tried this?  Do you have any hints on how to diagnose 
what is going wrong?


Once I switched the "Chipset" setting to "ICH9", it found the drive, and 
I was able to continue.


Also, I modified the config to use wired networking (using a virtual 
Intel-based NIC), but I couldn't think of a way to easily test whether 
it is working or not.  (I'm sure I'm missing something obvious here, but 
I appreciate any hints.)


Unfortunately, I am not having luck with the network card.  I have tried 
all of the available Adapter Types, but none of them work.  I get 
messages in "/report/log" like these:


For Intel 82540EM:

[runtimne -> nic -> nic] Found 00:3.0 8086:100e (rev 02) IRQ 09
[runtimne -> nic -> nic] using driver 82540em
[runtimne -> nic -> nic] PCI BIOS has not enabled device 00:3.0! 
Updating PCI command 0003->0007
[drivers -> platform_drv] Error: runtimne -> nic -> nic: assignment of 
PCI device 0:3.0 failed phys=0xdc018000 virt=0x1000
[runtimne -> nic -> nic] PCI device 00:3.0 latency timer is unreasonably 
low at 0. Setting to 32.

[runtimne -> nic -> nic] bus_addr = e200 len = 2
...
[runtimne -> nic -> nic] snprintf not implemented
[drivers -> platform_drv] IRQ override 9->9 trigger mode: LEVEL, 
polarity: LOW

[drivers -> platform_drv] 0:3.0 uses IRQ, vector 0x9, non-maskable


For Intel 82543GC (all the same, except):

[runtimne -> nic -> nic] Found 00:3.0 8086:100e (rev 02) IRQ 09
[runtimne -> nic -> nic] using driver 82543gc


For Intel 82545EM (all the same, except):

[runtimne -> nic -> nic] Found 00:11.0 8086:100f (rev 02) IRQ 0b
[runtimne -> nic -> nic] using driver 82545em
[runtimne -> nic -> nic] PCI BIOS has not enabled device 00:11.0! 
Updating PCI command 0003->0007
[drivers -> platform_drv] Error: runtimne -> nic -> nic: assignment of 
PCI device 0:11.0 failed phys=0xdc088000 virt=0x1000

...
[runtimne -> nic -> nic] bus_addr = e242 len = 2
...
[drivers -> platform_drv] 0:11.0 uses IRQ, vector 0xb, non-maskable


FWIW, VirtualBox doesn't seem to give any way of tweaking things at the 
IRQ level.


This is closer to the hardware than I usually get these days.  Is there 
anything I can try?


 Thanks!
--
  John J. Karcher
  devu...@alternateapproach.com

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Sculpt EA On VirtualBox

[John J. Karcher]

Hello,

I have downloaded, built, and begun testing Sculpt EA, following the 
instructions linked to in the 18.02 release notes.  Everything went well 
during the build process.


Since I primarily run everything I do in VirtualBox (and don't have the 
recommended hardware anyway), I created a VM as close to the 
recommendations as possible.  It booted fine from a virtual CD, the 
system came up using the VESA framebuffer driver, the mouse worked, and 
I was able to follow the instructions until I got to the part about 
storage.  Unfortunately, there are no "" nodes under 
"" in "/report/drivers/block_devices".


I have the VM set up with a "SATA" controller and a single virtual 
drive.  Has anyone tried this?  Do you have any hints on how to diagnose 
what is going wrong?


Also, I modified the config to use wired networking (using a virtual 
Intel-based NIC), but I couldn't think of a way to easily test whether 
it is working or not.  (I'm sure I'm missing something obvious here, but 
I appreciate any hints.)


If I can get those two items working, it looks like I'll be in good shape!

 Thanks,
--
  John J. Karcher
  devu...@alternateapproach.com

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: Time For "Community Development" ML?

[John J. Karcher]

On 01/26/2018 03:24 AM, Norman Feske wrote:

Hi John,

On 25.01.2018 23:36, John J. Karcher wrote:

What do you think of setting up a new "Community Development" mailing
list, for discussion and coordination within the development community
that doesn't need the attention of the core developers?


as far as I am concerned, I am not really concerned. ;-)


Glad to hear it.  ;^)


I think that the current mailing list works quite for the following reasons:

* It provides a single point of reference for Genode-related questions
   and discussions, which avoids potential confusion of new community
   members. E.g., the forum we once offered at Sourceforge turned out
   to be a mistake. After all, the more communication channels we have,
   the more likely a topic might go unnoticed by one person or another.

* The traffic is at a manageable level, and the overall quality and tone
   of the postings is nice. I don't feel overwhelmed by it.

* Advanced technical discussion and coordination happens at the issue
   tracker anyway. So there is already a distinct communication channel
   in place for getting work done.


We're probably already at the stage now where it would be helpful, but I
think the "Year Of Sculpt" will generate much wider hobbyist interest
than in the past, which could disrupt the S/N ratio of this list.

(Eventually, an end-user oriented ML would probably be a good idea also,
but that should be a year or more away.)


That is a very good point. Once we get an influx of Sculpt users, a
community-organized forum targeted specifically to the needs and
questions of end users would be a welcome contribution! Are you
interested in building up and moderating such a community place?


Unfortunately, I don't have the time (or temperament) to take a 
leadership role.  But as I learn and gain experience, I will certainly 
share and contribute what I can.



We may also consider a sub-reddit as a place for chitchatting, which
wouldn't perfectly suit the mailing list or a discussion forum.

Cheers
Norman



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Time For "Community Development" ML?

[John J. Karcher]

Hi Norman,

What do you think of setting up a new "Community Development" mailing 
list, for discussion and coordination within the development community 
that doesn't need the attention of the core developers?


We're probably already at the stage now where it would be helpful, but I 
think the "Year Of Sculpt" will generate much wider hobbyist interest 
than in the past, which could disrupt the S/N ratio of this list.


(Eventually, an end-user oriented ML would probably be a good idea also, 
but that should be a year or more away.)


 Thanks!
--
  John J. Karcher
  devu...@alternateapproach.com

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: Roadmap 2018

[John J. Karcher]

On 01/17/2018 11:45 AM, Norman Feske wrote:

Hello,

I'd like thank everyone who participated in the road-map discussion for
2018!

I just updated Genode's official road map and hope that it adequately
captures the interests expressed during the discussion while staying
realistic:

   https://genode.org/about/road-map

Cheers
Norman


As a hobbyist, I hesitate to be among the first to respond, but I have 
to say that I am very excited about "The Year Of Sculpt".  I think it 
could be a real turning point when it comes to gaining a wider audience.


I can't wait to get my hands on Sculpt EA!

In the meantime, please be sure to let us know if you upload any 
materials and/or videos from the FOSDEM presentations.


Thanks!

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main

Re: Roadmap 2018

[John J. Karcher]

On 12/27/2017 11:15 AM, Christian Helmuth wrote:

Hello,

thanks Norman for making a start with the road-map discussion and also
for the review of 2017. Personally, I'm quite excited that Sculpt
enables me to work on Genode each day. This is pretty much different
from what I felt one year ago with a rather static scenario that I
used only once for a talk.

Looking at 2017 in hindsight, I did not invest much time into my
personal plan, which was a Genode-native multi-component email
workflow including IMAP, SMTP, local (maildir) storage, and mutt as
MUA. Nevertheless, I'll extend the plan to 2018 as Sculpt promises
less obstacles and features true dynamic Genode subsystems beside the
traditional Linux VM. The first natural step is the use of multiple
VMs tailored for dedicated purposes of daily work (speak development,
email, web browsing). This is not my personal ambition but the first
pieces were already put in place by others in the team. This scenario
helps to understand and solve the task of sharing data when splitting
the daily work into more fine-grained domains. Next, the email VM can
be replaced by the Genode email subsystem developed in parallel.


Chris, please keep us informed about your progress on the IMAP server 
scenario if you can.  I am also gradually making my systems more modular 
using VMs (although not quite as granular as what you describe above), 
and I am currently running an little IMAP server in a *nix VM, which I 
would love to replace with a Genode "appliance" VM.



Everyone, since I'm already chiming in, I might as well throw in this 
idea as food for thought:


It might be possible to piggyback on the work of the "postmarketOS" 
project ( https://postmarketos.org/ ) to get Genode running on a variety 
of smartphones with only one porting effort.  In a nutshell, they are 
isolating all the device-specific code into one file, in order to allow 
creating a single (highly customizable, Alpine Linux-based) OS image to 
run on all the supported devices.  I wonder if it's possible to create a 
Genode-on-Linux (ARM) scenario on this foundation.


In any case, thanks for the amazing work, and Happy New Year to everyone!

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
genode-main mailing list
genode-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/genode-main