June 2003 - Darwin-kernel - lists.apple.com

Re: Current Process
by Kevin Van Vechten 10 Jun '03

10 Jun '03

Hi Joey, What API contains currentProcess()? What are you hoping for currentProcess() to return? - Kevin On Tuesday, June 10, 2003, at 09:53 AM, Joseph Gabriel Echeverria wrote: How can I get the current process from inside a kernel extension? I've tried calling currentProcess(), but when that is included, the KEXT won't load becuase of unresolved symbols. Thanks, Joey Joseph G. Echeverria Department of Electrical and Computer Engineering Carnegie Mellon University _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Current Process
by Joseph Gabriel Echeverria 10 Jun '03

10 Jun '03

How can I get the current process from inside a kernel extension? I've tried calling currentProcess(), but when that is included, the KEXT won't load becuase of unresolved symbols. Thanks, Joey Joseph G. Echeverria Department of Electrical and Computer Engineering Carnegie Mellon University _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Current Directory of a Process
by Justin C. Walker 09 Jun '03

09 Jun '03

On Monday, June 9, 2003, at 02:52 PM, Robert Sandilands wrote: I have a struct proc * inside a sysent[xxx] call and from there want to get the current working directory of the process that initiated the action. I have looked through the kernel and the mailing list archives and not found any obvious pointers. Does anybody have any ideas? You can look at xnu/bsd/vfs/vfs_syscalls:chdir() to see how the various structures are stitched together. Is this what you are looking for? Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large * Institute for General Semantics | Some people have a mental | horizon of radius zero, and | call it their point of view. | -- David Hilbert *--------------------------------------*-------------------------------* _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Current Directory of a Process
by Robert Sandilands 09 Jun '03

09 Jun '03

I have a struct proc * inside a sysent[xxx] call and from there want to get the current working directory of the process that initiated the action. I have looked through the kernel and the mailing list archives and not found any obvious pointers. Does anybody have any ideas? Robert Sandilands --------------------------------------------------------------------- #include http://robert.rsa3.com/disclaimer.html _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Having problems with .gdbinit kernel macros...
by Shawn Erickson 07 Jun '03

07 Jun '03

On Friday, June 6, 2003, at 09:35 AM, Garth Cummings wrote: Hi Shawn, On Thursday, June 5, 2003, at 07:35 PM, Shawn Erickson wrote: On Wednesday, June 4, 2003, at 09:07 AM, Quinn wrote: At 9:56 -0700 1/6/03, Shawn Erickson wrote: On Sunday, June 1, 2003, at 01:56 AM, Quinn wrote: At 13:57 -0700 31/5/03, Shawn Erickson wrote: Or do I need to have a debug/symboled build of the kernel for them to work? You need to have kernel symbols for the kernel macros to work. A debug kernel is not required. So a symboled kernel or just the symbols (what their is in mach.sym)? AFAIK there are two sources of kernel symbols. 1. If you build the kernel yourself (using the instructions from "IMOSX:Kernel Programming"), the result is a "mach_kernel" file, which you install on the victim machine, and a "mach_kernel.sys" file, which is the kernel symbols which you load into GDB on the build machine using "add-symbol-file". 2. When Apple builds Mac OS X ourselves, we build and archive this file. Currently we do not have a formal process for distributing these symbol files, but we're working on it. How about informally firing me a copy of the symbols for 10.2.6? Does that count as a 3rd source? It's about 13MB zipped. Too big to email. We're very close (within days) of implementing option 2. Days! ...and I just built the kernel I needed. ;-) _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Having problems with .gdbinit kernel macros...
by Garth Cummings 06 Jun '03

06 Jun '03

Hi Shawn, On Thursday, June 5, 2003, at 07:35 PM, Shawn Erickson wrote: On Wednesday, June 4, 2003, at 09:07 AM, Quinn wrote: At 9:56 -0700 1/6/03, Shawn Erickson wrote: On Sunday, June 1, 2003, at 01:56 AM, Quinn wrote: At 13:57 -0700 31/5/03, Shawn Erickson wrote: Or do I need to have a debug/symboled build of the kernel for them to work? You need to have kernel symbols for the kernel macros to work. A debug kernel is not required. So a symboled kernel or just the symbols (what their is in mach.sym)? AFAIK there are two sources of kernel symbols. 1. If you build the kernel yourself (using the instructions from "IMOSX:Kernel Programming"), the result is a "mach_kernel" file, which you install on the victim machine, and a "mach_kernel.sys" file, which is the kernel symbols which you load into GDB on the build machine using "add-symbol-file". 2. When Apple builds Mac OS X ourselves, we build and archive this file. Currently we do not have a formal process for distributing these symbol files, but we're working on it. How about informally firing me a copy of the symbols for 10.2.6? Does that count as a 3rd source? It's about 13MB zipped. Too big to email. We're very close (within days) of implementing option 2. (slowly convincing myself to get off my a$$ and build my own kernel...) -Shawn Thanks, --gc _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Socket NKE performance
by Stéphane Sudre 06 Jun '03

06 Jun '03

On jeudi, juin 5, 2003, at 19:35 Europe/Paris, Joshua Graessley wrote: On Thursday, June 5, 2003, at 7:33, Stiphane Sudre wrote: I'm testing what kind of performance I can get from a Socket NKE communication from a Kernel Extension to a user land application. I set the recv_space of the socket to be 224 KB in the Kernel In the application, I'm using a CFSocket to read the data when available. To communicate with the application I'm sbappending 32 Bytes data using the same kind of code that the NKE tcpdumper sample code. My unrealistic dream is to reach a peak of 1 MB/s. But 300 KB/s would already be fair. >From the test I'm making, the performance can just reach 25 KB/s. Is this not a bit weak compared to what should be expected? Is there a reason you're only writing 32 bytes at a time? Because I have to as the function may not be called anew for a while. And the "message" I need to send is 32 bytes long. Are you calling sorwakeup after each time that you write 32 bytes? Yes. I may need to clarify where I'm getting the 25 KB/s result from. Before the sbappend, sorwakeup calls, I have a test like this: freeSpace=sbspace(&_NKESocket->so_rcv); if (freeSpace < (32*sizeof(char)) { printf("No more space. (only %ld byte(s) available)\n",freeSpace); return; } ... p = m->m_data; p = (char *)(((int)p+3)&(~0x3)); m->m_data = (caddr_t)p; bcopy(&tData, mtod(m, caddr_t), (32*sizeof(char))); m->m_len = (32*sizeof(char)); m->m_flags |= M_EOR; sbappend(&_NKESocket->so_rcv, m); sorwakeup(_NKESocket); To simplify, this procedure may get call for every networking packet received or sent. To test the performance of this, from another machine, I'm getting some html files at a rate of 100 per seconds (using http_load). This leads to maybe more than 1000 packets being sent from my machine per second. What I'm seeing in the user land is that I'm only receiving 700-800 'messages' per second from the Kernel and that in the same time I'm seeing many "No more space. (only 0 byte(s) available)" in system.log per second. So doing a quick and may be stupid computing this gave me: 800 * 32 bytes /s 25,600 Bytes/s What I don't understand is since i have a so_rcv of 224 KB, I would have thought that I could get more than 25 KB/s. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Socket NKE performance
by Stéphane Sudre 06 Jun '03

06 Jun '03

On jeudi, juin 5, 2003, at 18:36 Europe/Paris, Justin C. Walker wrote: On Thursday, June 5, 2003, at 07:33 AM, Stiphane Sudre wrote: I'm testing what kind of performance I can get from a Socket NKE communication from a Kernel Extension to a user land application. I set the recv_space of the socket to be 224 KB in the Kernel In the application, I'm using a CFSocket to read the data when available. To communicate with the application I'm sbappending 32 Bytes data using the same kind of code that the NKE tcpdumper sample code. My unrealistic dream is to reach a peak of 1 MB/s. But 300 KB/s would already be fair. >From the test I'm making, the performance can just reach 25 KB/s. Is this not a bit weak compared to what should be expected? Yup, it's quite low. If you are just in a loop, sbappending 32 bytes at a time, I don't know that you'll get high performance, and you will certainly churn up a lot of CPU cycles no matter what. What are you trying to achieve? Basically doing what tcplogger is doing without having to use a waiting list when there is no more so_rcv space available. This is the reason why I increased the so_rcv space for the NKE socket. I want to do this for 3 reasons: 1) I can't be sure that when one event is to be sent and there is no memory available at this instant, there will be another opportunity to send this event. 2) I want the smallest delay between the event and the report of the event 3) I'm afraid of the amount of memory I may have to allocate in the Kernel if I implement a waiting list. But I may be wrong as I'm always thinking that 1 KB is a lot and one i++ instruction is hogging the CPU. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Having problems with .gdbinit kernel macros...
by Shawn Erickson 06 Jun '03

06 Jun '03

On Wednesday, June 4, 2003, at 09:07 AM, Quinn wrote: At 9:56 -0700 1/6/03, Shawn Erickson wrote: On Sunday, June 1, 2003, at 01:56 AM, Quinn wrote: At 13:57 -0700 31/5/03, Shawn Erickson wrote: Or do I need to have a debug/symboled build of the kernel for them to work? You need to have kernel symbols for the kernel macros to work. A debug kernel is not required. So a symboled kernel or just the symbols (what their is in mach.sym)? AFAIK there are two sources of kernel symbols. 1. If you build the kernel yourself (using the instructions from "IMOSX:Kernel Programming"), the result is a "mach_kernel" file, which you install on the victim machine, and a "mach_kernel.sys" file, which is the kernel symbols which you load into GDB on the build machine using "add-symbol-file". 2. When Apple builds Mac OS X ourselves, we build and archive this file. Currently we do not have a formal process for distributing these symbol files, but we're working on it. How about informally firing me a copy of the symbols for 10.2.6? Does that count as a 3rd source? (slowly convincing myself to get off my a$$ and build my own kernel...) -Shawn _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Socket NKE performance
by Joshua Graessley 05 Jun '03

05 Jun '03

On Thursday, June 5, 2003, at 7:33, Stiphane Sudre wrote: I'm testing what kind of performance I can get from a Socket NKE communication from a Kernel Extension to a user land application. I set the recv_space of the socket to be 224 KB in the Kernel In the application, I'm using a CFSocket to read the data when available. To communicate with the application I'm sbappending 32 Bytes data using the same kind of code that the NKE tcpdumper sample code. My unrealistic dream is to reach a peak of 1 MB/s. But 300 KB/s would already be fair. >From the test I'm making, the performance can just reach 25 KB/s. Is this not a bit weak compared to what should be expected? Is there a reason you're only writing 32 bytes at a time? Are you calling sorwakeup after each time that you write 32 bytes? -josh _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0