Re: A niche for the Hurd - multi-core systems

n any case, IMHO we have much more pressing issues than Mach's > shortcomings presently, and thus I don't consider work on new > microkernels the highest priority right now... (Definitely interesting > though.) but I think at least some bugs in Mach should be fixed. Zheng Da

Re: looking for the solution of rootless subhurd

tricky, > though.) I doubt it can work. We can inject the send right to the process when we create it, but the problem is how the process know the name of the port (the name of port, I mean, is the integer value). It might already be out of topic because we don't need to deal with it if the possible solution I found above works. Zheng Da

Re: looking for the solution of rootless subhurd

ent to do some experiments, but somehow the program in one subhurd can crash the other subhurd. Isn't it very strange? Since I think subhurd as a virtual machine, I hope it can give more control of resource usage to the user. For example, limit its memory usage (It's the only example I can think of now:-). Now subhurd can only use the devices specified by the user with "-m" option when it run by the normal user. Zheng Da

Re: the virtual device management in eth-multiplexer

rking as soon as possible:-) I have already implemented all RPCs related to the task port, and now need to see how it works. Zheng Da

Re: the virtual device management in eth-multiplexer

ntly storing translator hierarchies. > > I would like to talk about that. Can you make it to the meeting this > friday? (It is rather urgent, as I'm considering making a GSoC task out > of it...) > I think it will be OK. Let's meet at 7:00pm this Friday. Zheng Da

Re: the virtual device management in eth-multiplexer

On Fri, Mar 6, 2009 at 5:08 AM, wrote: > Hi, > > On Thu, Mar 05, 2009 at 04:10:13PM +0000, Zheng Da wrote: > > On Thu, Mar 5, 2009 at 2:23 AM, wrote: > > > > Can you make it to the meeting this friday? (It is rather urgent, as > > > I'm considering mak

Re: The problems for the rootless subhurd

t tasks but I cannot tell where are the RPCs from, because a task can send the RPC on the behalf of another task. Zheng Da

Re: GSoC IRC meeting

Hi, I will go back to China at the end of this month, so I won't join the meeting any more after May. Zheng Da On Wed, May 6, 2009 at 4:47 PM, Sergiu Ivanov wrote: > Hello, > > On Wed, May 6, 2009 at 3:31 PM,   wrote: >> I learned recently that I will be away for the week

[patch #6830] Trace all tasks and threads

URL: Summary: Trace all tasks and threads Project: The GNU Hurd Submitted by: zhengda Submitted on: Sat 16 May 2009 05:24:05 PM GMT Category: other Hurd Priority: 5 -

[patch #6833] rpctrace: handle mach_port_insert_right()

d know which task has it. Again, mach_port_extract_right() is a very difficult case to handle. Some other problems are left in the patch. For example, how to destroy the wrapper? Could anyone give me some suggestions? Thank you, Zheng Da ___

[PATCH] fully enable rpctrace to trace multitask programs.

o effect i486-gnu-gcc: no input files mig: fatal: no SubSystem declaration The patch is below. Comments and tests are very welcome. Thank you, Zheng Da 2009-05-31 Zheng Da *rpctrace.c: include sys/mman.h. (UNKNOWN_TASK): New macro. (UNKNOWN_NAME): Likewise. (tr

[patch #6844] fix a bug in init: initialize 'flags' in reboot_system

URL: Summary: fix a bug in init: initialize 'flags' in reboot_system Project: The GNU Hurd Submitted by: zhengda Submitted on: Tue 02 Jun 2009 08:47:27 AM GMT Category: other Hurd

[bug #26751] creating memory objects through the proxy defpager can fail

Follow-up Comment #1, bug #26751 (project hurd): I had a problem the first time I used proxy defpager. I don't know if we have the same problem. The attached patch might help. At least it solves my problem. (file #18238) ___ Additional It

[bug #26788] GNUMach crashes after rpctrace exits unexpectedly

U Mach Severity: 3 - Normal Priority: 5 - Normal Item Group: None Status: None Privacy: Public Assigned to: None Originator Name: Zheng Da Originator Email: zhengda1...@gmail.com Open/

Re: [PATCH] fully enable rpctrace to trace multitask programs.

ome places. The ChangeLog is missing. I will add it later in case that there will be another big change in the patch. Zheng Da diff --git a/utils/rpctrace.c b/utils/rpctrace.c index d80f41d..6093c03 100644 --- a/utils/rpctrace.c +++ b/utils/rpctrace.c @@ -59,6 +59,8 @@ static const struct argp_opti

[bug #20612] rpctrace: heisenbug

Follow-up Comment #1, bug #20612 (project hurd): The patch in the attachment can fix the bug. (file #18272) ___ Additional Item Attachment: File name: rpctrace1.diff Size:0 KB

[patch #6851] fix a bug in BPF

URL: Summary: fix a bug in BPF Project: The GNU Hurd Submitted by: zhengda Submitted on: Sun 21 Jun 2009 01:36:19 AM GMT Category: GNU Mach Priority: 5 - Normal

Re: Setting up eth-multiplexer in QEMU

1. patch #6619: pfinet uses the virtual interface 2. patch #6620: pfinet changes its filter rules with its IP address 3. patch #6621: pfinet sets the mach device into the promiscuous mode. 4. patch #6622: pfinet uses the BPF filter It should be very helpful for us to locate the problematic patch. Zheng Da

Re: plan to work on user-level device drivers

CMU's single-server UNIX emulator? Where can I >> download it? > > Are you sure it's from CMU?... I thought it was from CMU. I am not sure. Zheng Da

[bug #25054] Kernel panic with eth-multiplexer

Follow-up Comment #1, bug #25054 (project hurd): The kernel panic was caused by a bug in eth-multiplexer, which didn't deallocate the out-of-line data in ds_device_write(). The problem doesn't exist any more. ___ Reply to this item at:

Re: plan to work on user-level device drivers

tions enforce the rule that papers must not be > published privately, and they might not like the paper being in the open, so > better be sure it's allowed... > No, I don't see any messages about it. The one I get seems to be a scan copy. Maybe I shouldn't upload it. Zheng Da

[patch #6851] fix a bug in BPF

Follow-up Comment #2, patch #6851 (project hurd): The packet delivered from gnumach should have the packet header. NPF returns the packet_header and pfinet always assumes that the packet from gnumach has the packet_header. For your first question, the return value of net_do_filter is boolean, i.

[patch #6851] fix a bug in BPF

Follow-up Comment #4, patch #6851 (project hurd): The only user of packet_header that I can see is NPF. It requires more data copy in order to remove the packet_header before delivering the packet to the user space, I guess. but I don't understand the implementation of NPF.

[patch #6967] fix a bug in vm_page_grab_contiguous_pages

Update of patch #6967 (project hurd): Wiki-like text discussion box: => Sorry for that. Vim usually did the wrapping job for me so I didn't pay much attention on it. I will be more careful next time. ___ Reply to t

Re: Hurd DDE test

could clone the repository in my home directory on flubber. If you have a flubber account, you might be able to do it. Best regards, Zheng Da On Wed, Dec 16, 2009 at 10:20 AM, Shakthi Kannan wrote: > Hi Zheng, > > Can you please let me know what you have done so far w.r.t. DDE and &g

Re: GNU Mach: disabling all network device drivers

17 I don't know whether the error is caused by my change in gnumach or the error exists in gnumach itself. It's probably not very easy to check as the error occurs when a user-level driver runs. Best regards, Zheng Da

Re: GNU Mach: disabling all network device drivers

Hello, On Sun, Jun 6, 2010 at 5:57 PM, Samuel Thibault wrote: > Hello, > > Zheng Da, le Thu 13 May 2010 08:59:59 -0400, a écrit : >> I boot hurd with the new kernel and run the user-level network drivers >> but got a kernel >> general protection trap. I tried several

Re: GNU Mach: disabling all network device drivers

Hi, On Mon, Jun 7, 2010 at 4:20 AM, Samuel Thibault wrote: > Zheng Da, le Mon 07 Jun 2010 01:40:26 -0400, a écrit : >> > examine 2f0094 >> db> examine 0x1973e0 >>                 13f163 > > Where does this point to in the source code? (null):~# addr2line -e /

Re: GNU Mach: disabling all network device drivers

/* program PICs with new mask */ sti /* enable interrupts */ nop movl%edx,%eax /* return previous ipl */ ret So which instruction triggers the trap? nop or movl? Zheng Da

Re: GSoC project about virtualization using Hurd mechanisms

Hi, I'll try to do it, but it will take a little long time because I have a lot of projects from my school these days. So could you give me some more advices to help me get into the project faster? For example, which article or which part of code should I read first? Best, Zheng Da On Fri

GSoC: the plan for the project network virtualization

much work it could be? And how do I start? Best, Zheng Da

Some questions about the translator that helps open the device

evnode, trivdev, devconn. any ideas? Best regards, Zheng Da

Re: GSoC projects

se git or else for it instead of > CVS, whatever people prefer). If it has to be integrated into the main > Hurd tree, then we have to make up some rules about how to do this. a) I'm Zheng Da, and I'm working on Network virtualization. The idea is to run multiple pfinet servers on

Re: The patch of gnumach to enable the user to set the network device into the promiscuous mode

I copied the code from http://perso.hurdfr.org/mmenal/promisc.patch. I think only the root should be able to set the flags. Zheng Da On Fri, Aug 8, 2008 at 9:33 PM, Samuel Thibault < [EMAIL PROTECTED]> wrote: > Hello, > > zhengda, le Fri 08 Aug 2008 06:44:16 +0200, a écrit : >

[patch #6618] Enable boot to use the device file created by devnode.

URL: Summary: Enable boot to use the device file created by devnode. Project: The GNU Hurd Submitted by: zhengda Submitted on: Fri 29 Aug 2008 01:54:25 PM GMT Category: other Hurd

[patch #6619] pfinet uses the virtual interface

URL: Summary: pfinet uses the virtual interface Project: The GNU Hurd Submitted by: zhengda Submitted on: Fri 29 Aug 2008 01:56:52 PM GMT Category: other Hurd Priority

[patch #6620] pfinet changes its filter rules with its IP address

URL: Summary: pfinet changes its filter rules with its IP address Project: The GNU Hurd Submitted by: zhengda Submitted on: Fri 29 Aug 2008 01:59:02 PM GMT Category: other Hurd

[patch #6621] pfinet sets the mach device into the promiscuous mode.

URL: Summary: pfinet sets the mach device into the promiscuous mode. Project: The GNU Hurd Submitted by: zhengda Submitted on: Fri 29 Aug 2008 02:03:26 PM GMT Category: other Hurd

[patch #6622] pfinet uses the BPF filter

URL: Summary: pfinet uses the BPF filter Project: The GNU Hurd Submitted by: zhengda Submitted on: Fri 29 Aug 2008 11:51:31 PM GMT Category: other Hurd Priority: 5 - N

[patch #6624] provides the interface for the program to set the kernel device into promiscuous mode.

URL: Summary: provides the interface for the program to set the kernel device into promiscuous mode. Project: The GNU Hurd Submitted by: zhengda Submitted on: Sun 31 Aug 2008 02:41:26 PM GMT

[patch #6624] gnumach provides the interface for the program to set the kernel device into promiscuous mode.

Update of patch #6624 (project hurd): Summary: provides the interface for the program to set the kernel device into promiscuous mode. => gnumach provides the interface for the program to set the kernel device into promiscuous mode. __

[patch #6625] glibc privdes environment variables to ovrride the default pfinet server.

Update of patch #6625 (project hurd): Summary: privdes environment variables to ovrride the default pfinet server. => glibc privdes environment variables to ovrride the default pfinet server. ___ Reply to this item at:

[patch #6625] privdes environment variables to ovrride the default pfinet server.

URL: Summary: privdes environment variables to ovrride the default pfinet server. Project: The GNU Hurd Submitted by: zhengda Submitted on: Sun 31 Aug 2008 02:42:29 PM GMT Category: G