On 10-5-11 上午9:35, Da Zheng wrote:
> Hello,
>
> I'm testing the porting of DDE Linux26 with different types of NIC drivers on
> VMWare and qemu. I tested pcnet32, ne2k-pci, rtl8139, e1000. pcnet32 works in
> both VMWare and qemu. Only qemu emulates ne2k-pci and rtl8139 cards and both DDE
> drivers work well in qemu.
>
> The odd thing occurs to e1000. VMWare and qemu both emulate e1000. DDE e1000
> doesn't work in VMWare but work quite well in qemu. I don't know why.
…
[View More]The bug in e1000 is finally fixed. It should be caused by the problem cache
coherency. After I made the DMA buffer cacheline-aligned, e1000 driver works
well in VMWare.
>
> All code (DDE Linux26 and modified gnumach) is in the dde branch of incubator
> and in master-user_level_drivers branch of the gnumach repository. I appreciate
> if someone can help me test the drivers in a real machine or test other NIC
> drivers. Instructions to build and run drivers can be found at
> http://www.bddebian.com:8888/~hurd-web/user/zhengda/
Zheng Da
[View Less]
I know that l4linux cpu always is allocated on 4th thread on linux server.
Also, I want to assign specific timeslice into the linux CPU thread to operate its linux kernel and its user task.
To do this, I wanted to be expected in one of linux can use CPU more than another linux. However, even though I assigned timeslice more (it can be done by adding schedule context), 2 linux servers is seemed to be using almost same the amount of CPU. I think that this is because of idle thread. If idle …
[View More]thread is executed, other linux can preempt the linux. Am I right?
also, I'm wondering whether user task consumes their own timeslice or donated timeslice from CPU thread.
Thank you
[View Less]
I've found that this code is not wrong. The problem was from IPC control in thread-ipc.c
in sys-ipc() function, the kernel check whether the ipc is next_period() or not, but the check function is wrong i think because the function returns always "FALSE".
That means that there is no next_period() ipc.
so I've changed the code in shared/entry_frame-ia32-ux.cpp line 110
return false;
-->
return (_ecx == 0x04000400 && _esi == 0);
I'm not sure that this code does not affect to the …
[View More]other ipc. However it is executing quite well.
Jungwoo Yang.
[View Less]
Hello,
In DDE Linux26, the interrupt is ignored when an interrupt line is disabled,
i.e., the handle_irq field of ddekit_irq_ctrl is 0.
In ne2k-pci, an example in DDE Linux26, when a packet is transmitted by the
driver, disable_irq_nosync_lockdep_irqsave() is called to disable irq. Then it
triggers a send and calls enable_irq_lockdep_irqrestore().
Now the problem is that after a packet is transmitted, the transmission-done
interrupt can be delivered before enable_irq_lockdep_irqrestore is …
[View More]called, so
the interrupt is ignored and the driver thinks the packet cannot be transmitted.
This situation happens when I run ne2k-pci in the Hurd porting of DDE Linux26.
disable_irq_nosync_lockdep_irqsave is exactly the same as disable_irq_nosync
when there isn't CONFIG_LOCKDEP. It seems to me that ne2k-pci shouldn't call it
in the first place when transmitting packets, but Linux drivers should be
correct. Then what goes wrong? Should the interrupt handler delays interrupts
instead of ignoring them silently when the interrupt line is disabled?
Best regards,
Zheng Da
[View Less]
Hi Adam,
My application uses Time Warped simulation library which uses standard c++
library. The time warped library uses a class called multimap which has a
default constructor in std c++ library while in uclibc++ there is no default
constructor. Hence to conform to uclibc++ I need to pass few arguments to
use the current constructor in uclibc++ but this screws up a bit. I can
modify multimap in uclibc++ with a default constructor but I am not sure
whether default constructor is omitted …
[View More]intentionally and has some side
effects. Also I need to do this change to all classes it is inheriting.
Thats why I thought is it possible to use std c++ library instead of
uclibc++. Can I know the version of std c++ library that is ported to
uclibc++? Thanks a lot for your comments.
Regards,
Karthik
[View Less]
Just so you know, I was a big person in information when it was
centered in my home city of NY.
Just as NYC's tech industry collapsed (mostly, I believe because of
the bombing of the WTC--which itself was a headquarters of free
software), I was completing the design criterion of the Thinman model,
which is an extension of Perl's CPAN to a VM that is open to the
Internet for both data and code. It is anti-app in that each object,
presumably a dataset, is, as it is evoked, its own app. Your …
[View More]friends
list, as a small example, becomes a phone book in a flexible browser,
when it is evoked in the context of a phone call through a model or
other connection.
I personally know David Korn, which is significant to me, since I
believe that his KSH, with its related commands and related shells
such as BASH, has provided the best work, and play, environment we
have. It has a little awkwardness (which is forgivable given it was a
proof of concept at its inception in the 1970s), and no OO version
Shell has been attempted. And needless to say, it lacked a visual
desktop!
Other than those few issues, I cannot help but believe his traditional
open shell still provides a model of the type of interactive framework
that we need to bring to L4 as the default interface.
I have since taken social science paths, and perhaps because of this,
I very much want to help work from the top down (as you are with your
ground-breaking port) to the lower levels, where everybody else is, so
that L4 growth can be what free software is really all about--our
system!
John van V.
On Sun, May 9, 2010 at 1:02 PM, Karthik Vadambacheri Manian
<karthikvm.uc(a)gmail.com> wrote:
> Thanks for the comments John. Its very encouraging!!
>
> Regards,
> Karthik
>
> On Sat, May 8, 2010 at 3:41 PM, John van V. <john.van.v(a)gmail.com> wrote:
>>
>> Great work, seriously!
>>
>> On Fri, May 7, 2010 at 12:17 PM, Karthik Vadambacheri Manian
>> <karthikvm.uc(a)gmail.com> wrote:
>> > Hi All,
>> >
>> > Currently I am trying to port a pingpong c++ application which uses time
>> > warped libraries. The time warped libraries use standard c++ library
>> > inturn.
>> > I tried to compile the pingpong application by modifying the makefiles
>> > to
>> > include the time warped headers. But I get lot of errors while compiling
>> > because during compilation the uclibc++ header files are used instead of
>> > standard c++ headers and they differ. Hence I tried to port the warped
>> > library functions for uclibc++ but it seems it needs a huge
>> > modification. Do
>> > modifying makefiles to use standard c++ header files instead of uclibc++
>> > will break other systems? Please let me know your comments.
>> >
>> > Thanks,
>> > Karthik.V.M.
>> > _______________________________________________
>> > l4-hackers mailing list
[View Less]