January 2003 - Darwin-kernel

Re: Writing Modules for Darwin
by Quinn 30 Jan '03

30 Jan '03

At 7:57 +0000 30/1/03, Jean-Matthieu Schaffhauser wrote: Is there any howto on Darwin modules programming available somewhere on the Internet. I'm totally new to kernel development and I have know idea where to start from. Any advice would be greatly appreciated. Depends what you mean by "modules". Under Darwin, kernel plug-ins are called "kernel extensions", or KEXTs. Most KEXTs done within I/O Kit, which in a C++ framework for device drivers. You can also write KEXTs that plug-in to the networking stack (network kernel extensions, or NKEs) or the file system. Finally, there are generic KEXTs. There's a lot of documentation about this stuff on the Apple developer web sit. <http://developer.apple.com/techpubs/macosx/Darwin/index.html> I/O Kit drivers are the best documented form of KEXTs because that's what we encourage developers to use. Apple does not (yet) give any binary compatibility promises for other forms of KEXTs. S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Technical Support * Networking, Communications, Hardware _______________________________________________ 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

Writing Modules for Darwin
by Jean-Matthieu Schaffhauser 30 Jan '03

30 Jan '03

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, this is my first post on the list, sorry if I go over something that's already been discussed. Is there any howto on Darwin modules programming available somewhere on the Internet. I'm totally new to kernel development and I have know idea where to start from. Any advice would be greatly appreciated. Cheers, Jean-Matthieu -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.7 (GNU/Linux) iD8DBQE+ONr5FPfU0aAWdgQRAo/6AJwO/5xUxQLSFf2wUQpYxUq8gEKwggCgl9+U YO7I0HBecHJW8R7/BPo9qDM= =VLfB -----END PGP SIGNATURE----- _______________________________________________ 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: Kernel -> user boundary
by Marek Kozubal 29 Jan '03

29 Jan '03

On Mon, 27 Jan 2003, Godfrey van der Linden wrote: > >IS there anything I can do or use to initiate a > >kernel -> user crossing without user parts help? > > Not really. Do you have a user land app that you are talking to? If > so it is pretty easy to emulate a kernel initiated event. > > What do you want to happen in user land when the kernel hits something. I also need to be able to do this sort of communication. Though I'm fine with having some setup at the beginning in the user -> kernel direction. What my kext needs to do is send the user land process some information, and then wait for a reply back from the user land process with an answer to the query (the user daemon is doing some IOKit stuff during the wait time). The reply time is on the order of 15 seconds so I need to be able to release the kernel funnels while I'm waiting for a reply. This is all in a file system kext. In a previous conversation I had w/someone, it was suggested that socket's might be the way to go for this communication but perhaps there is something more appropriate. ioctl and sysctl dont feel appropriate which is whats "suggested" for file system communcaition. I'm not familiar w/the mach ports and the like, would these be better? The amount of data being communicated is variable length but on the order of 3 to 100 bytes or so. So I guess, which communications API should I start rummaging though to implement this? And what kind of setup from the user daemon will be required to the kext if the kext itself cannot initiate the connection? --- Marek Kozubal marek(a)portents.com _______________________________________________ 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

system call reflection
by Joshua LeVasseur 28 Jan '03

28 Jan '03

Does Darwin provide a kernel mechanism to reflect system calls to a third-party task? The solutions I've investigated: - ptrace(), but no functionality similar to Linux's PTRACE_SYSCALL - task_set_exception_ports() with EXC_MASK_SYSCALL : looking through the Darwin source code, looks as if an exception is sent to the exception port only for invalid system call numbers. - task_set_emulation_vector() : looks to be unused in Darwin. I could imagine it being useful for x86, where the INT instruction can specify an immediate interrupt vector entry point, and thus provide non-overlapping system call support. But for PowerPC, the only system call info is contained in R0, which makes it difficult to differentiate between a legitimate Darwin system call and an emulated system call (plus how would you capture an emulated sys call and then call a real Darwin system call without recursively jumping to the emulated exception vector again?). - I looked through the system call invocation path in the kernel for hooks that might reflect the system call. I just hope to avoid a user-space solution :) Thanks for any help, 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

Re: Kernel -> user boundary
by George André 28 Jan '03

28 Jan '03

Umesh Vaishampayan wrote: Sigh. I'm not even going to ask why :-( File renaming depending on file contents. Customer requirement. Therefore I can only do this at VOP_CLOSE time or later, since that's when I know about the file contents. Hope you grab all the necessary vnode locks in correct locking order, yes? Yeh, I guess so, I've took the kernel implementation of rename() syscall and changed it to fit my needs and it works correctly (called from inside my VOP_CLOSE. The other one of my solutions was to defer this renaming till a VOP_GETATTR was received on "", but my rename-wannabe panicked inside namei() (lockmgr: locking against myself) Obviously a recursive lock which I spent time debugging, but couldn't figure out exactly which one it was. That's why I wanted to try a user process approach quickly to see if it's feasible that way before I come back to this. Yes you do get called. You should still get a VOP_LOOKUP() call chmod() -> namei() -> lookup() -> VOP_LOOKUP(). Humm, yes, you're absolutely right, I missed this one. How long do you plan to "defer" this task? How do you plan to avoid that users process racing the Finder and renaming the file before Finder gets around calling chmod? How do you know others are really done using this file? Not too long so that user starts messing with the file and not too short so that the user process renames the file before the Finder calls chmod/chown. Also, all operations would be blocked till the rename was done.. (except VOP_REMOVE, LOOKUP and related..) Why can't you do what you want to do in *your* VOP_INACTIVE? Here you lost me, honestly. As far as I know, VOP_INACTIVE is called when a vnode is not active, but even so, are you proposing to fake vnodes for a period of time or something like that? PLease elucidate... Why shared memory between kernel and user process? Is this going to be used to move lots of data? Seems like wrong things to do [at least to me]. But then again without knowing what you are trying to do, it's impossible to say. As said, the user process approach is an alternative to the in-kernel VOP_GETATTR on filename "". Both approaches are kind of shaky IMHO, so please, if you know of a better solution (in-kernel or not) I would be glad to hear about it. All I want to do is rename a vnode at or after VOP_CLOSE... (I do know the name/path of the file to rename). Thanks, George. --Umesh Godfrey van der Linden wrote: At 21:09 +0100 03-1-27, George Andri wrote: Hi there, I am aware of user -> kernel boundary crossing, like Mach IPC, RPC, syscall, sysctl and ioctl. But I'm interested ina kernel -> user *initiated* connection, is there anything outthere? I have a kext that needs to initiate communication with an user process. I can alternatively let the user process initiate the connection in one of the above ways and then pass a data pool pointer, but that would be a lot harder. IS there anything I can do or use to initiate a kernel -> user crossing without user parts help? Not really. Do you have a user land app that you are talking to? If so it is pretty easy to emulate a kernel initiated event. What do you want to happen in user land when the kernel hits something. I guess what I'm asking is what problem are you actually trying to solve. Godfrey van der Linden _______________________________________________ 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. -- Umesh Vaishampayan Mac OS X - Kernel _______________________________________________ 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: Which library/Framework should i include for definition of host_s tatistics() function?
by Quinn 28 Jan '03

28 Jan '03

At 19:41 +0530 28/1/03, Ashish Kumar wrote: I used functions like - host_page_size() and mach_host_self() System.framework from /System/Library/Frameworks S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Technical Support * Networking, Communications, Hardware _______________________________________________ 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

Which library/Framework should i include for definition of host_s tatistics() function?
by Ashish Kumar 28 Jan '03

28 Jan '03

Hi all, I am not sure if its the correct forum to put my question. May be someone can still help. Which library/framework should i include in my codewarrior project to get the definition of these functions? I used functions like - host_page_size() and mach_host_self() The declarations are in <mach/mach_host.h> I am getting linker error. I tried to add few frameworks and libraries but have not been successful yet. Thanks, Ashish -----Original Message----- From: Matthias Ringwald [mailto:mringwal@inf.ethz.ch] Sent: Tuesday, January 28, 2003 3:45 PM To: darwin-kernel(a)lists.apple.com Subject: I2C usage and "how to wake up from Infrared" HI there I just finished porting the linux infrared control (LIRC) project to darwin. As we all know, there is no real I/O on macs so I conected the Infrared Receiver to the sound input of my sawtooth g4 that at least has this input. Lirc can decode most IR controls. (www.lirc.org www.inf.ethz.ch/~mringwal/lirc/ )" To really use my mac as a multimedia center, I need to have it in deep sleep and want to wake it up via the IR remote. When I press a button on the IR remote, I could use a capacitor and provide a normal TTL signal that indicates "mac should be running" Any simple ideas, what I could use? So far, I've seen the wake on modem call (although I need my internal modem) and wake on ethernet. what does this mean technically ? One solution would be to use a USB mouse or USB keyboard and connect the IR signal to the any button. this doesn't look optimal and if the mac is running I would have to deal with to much input... and would cost me a USB mouse/keyboard. Any other option? Next, and more interesting: has anybody ever tried to use the I2C busses inside the mac to connect external devices ? most micro-controllers have build-in I2C drivers or at least i2C libraries are available whereas not much MUCs does have USB support and then a lot of usb code would be required both on the mac as on the mcu. I found the linux lm_sensor project: http://www2.lm-sensors.nu/~lm78/index.html that descibres an easy way to get the 4 wires for the I2C bus from a memory dimm and also all necessary drivers But next, how do communicate with a microcontroller? I've seen the I2C classes in darwin, but with only the source code and the short api docu, its quite hard to get a bigger picture. I'm thinking about adding IrDA support to my mac, use the MCU for sending and receiving of IR signals, provide a serial port and e.g. 16 free programable digital I/O lines never mind ADC (thinking about an Atmel ATmega128).. The IrDA option would be very interesting inside my iBook, too... ok. any info ? thanks for reading this long mail, matthias ringwald eth z|rich _______________________________________________ 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

I2C usage and "how to wake up from Infrared"
by Matthias Ringwald 28 Jan '03

28 Jan '03

HI there I just finished porting the linux infrared control (LIRC) project to darwin. As we all know, there is no real I/O on macs so I conected the Infrared Receiver to the sound input of my sawtooth g4 that at least has this input. Lirc can decode most IR controls. (www.lirc.org www.inf.ethz.ch/~mringwal/lirc/ )" To really use my mac as a multimedia center, I need to have it in deep sleep and want to wake it up via the IR remote. When I press a button on the IR remote, I could use a capacitor and provide a normal TTL signal that indicates "mac should be running" Any simple ideas, what I could use? So far, I've seen the wake on modem call (although I need my internal modem) and wake on ethernet. what does this mean technically ? One solution would be to use a USB mouse or USB keyboard and connect the IR signal to the any button. this doesn't look optimal and if the mac is running I would have to deal with to much input... and would cost me a USB mouse/keyboard. Any other option? Next, and more interesting: has anybody ever tried to use the I2C busses inside the mac to connect external devices ? most micro-controllers have build-in I2C drivers or at least i2C libraries are available whereas not much MUCs does have USB support and then a lot of usb code would be required both on the mac as on the mcu. I found the linux lm_sensor project: http://www2.lm-sensors.nu/~lm78/index.html that descibres an easy way to get the 4 wires for the I2C bus from a memory dimm and also all necessary drivers But next, how do communicate with a microcontroller? I've seen the I2C classes in darwin, but with only the source code and the short api docu, its quite hard to get a bigger picture. I'm thinking about adding IrDA support to my mac, use the MCU for sending and receiving of IR signals, provide a serial port and e.g. 16 free programable digital I/O lines never mind ADC (thinking about an Atmel ATmega128).. The IrDA option would be very interesting inside my iBook, too... ok. any info ? thanks for reading this long mail, matthias ringwald eth z|rich _______________________________________________ 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: Kernel -> user boundary
by Umesh Vaishampayan 28 Jan '03

28 Jan '03

On Monday, January 27, 2003, at 03:30 PM, George Andri wrote: Hi Godfrey, My problem starts inside msdosfs.kext, which is my kext. I've changed this kext to rename certain files when created. Sigh. I'm not even going to ask why :-( (when the filesystem receives close op). Using NDINIT/namei/VOP_RENAME works ok Hope you grab all the necessary vnode locks in correct locking order, yes? and does what it's supposed to, but the problems come from the following chown/chmod syscalls from the Finder. Finder does that last on a file drag & copy operation. Problem is that since I've changed the name, chown and chmod fail inside their kernel implementations and never get down to me... Yes you do get called. You should still get a VOP_LOOKUP() call chmod() -> namei() -> lookup() -> VOP_LOOKUP(). Thus, our friend the Finder puts up a nice dialog about not finding the files. One of my solutions is to send the path name to the user process and it will do the rename itself and thus defer this task. How long do you plan to "defer" this task? How do you plan to avoid that users process racing the Finder and renaming the file before Finder gets around calling chmod? How do you know others are really done using this file? Why can't you do what you want to do in *your* VOP_INACTIVE? I'm currently working on this approach: the user proc allocates mem with vm_allocate(), the it uses mach_make_memory_entry_64() and then sends the ipc_port_t to the kext thru mount args. (user proc is actually spawned by mount_msdos). This allocated memory is then vm_mapped by > the kext and there you go... memory sharing. Is this approach something that will work, or am I out chasing rabbits? Thanks, George. Why shared memory between kernel and user process? Is this going to be used to move lots of data? Seems like wrong things to do [at least to me]. But then again without knowing what you are trying to do, it's impossible to say. --Umesh Godfrey van der Linden wrote: At 21:09 +0100 03-1-27, George Andri wrote: Hi there, I am aware of user -> kernel boundary crossing, like Mach IPC, RPC, syscall, sysctl and ioctl. But I'm interested ina kernel -> user *initiated* connection, is there anything outthere? I have a kext that needs to initiate communication with an user process. I can alternatively let the user process initiate the connection in one of the above ways and then pass a data pool pointer, but that would be a lot harder. IS there anything I can do or use to initiate a kernel -> user crossing without user parts help? Not really. Do you have a user land app that you are talking to? If so it is pretty easy to emulate a kernel initiated event. What do you want to happen in user land when the kernel hits something. I guess what I'm asking is what problem are you actually trying to solve. Godfrey van der Linden _______________________________________________ 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. -- Umesh Vaishampayan Mac OS X - Kernel _______________________________________________ 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

IOHIDSystem message recurring
by Paul J. Teeter 28 Jan '03

28 Jan '03

We keep seeing the same error in our syslog, it has coincided with 2 complete system crashes: Jan 27 16:16:52 www mach_kernel: IOHIDSystem: postEvent LLEventQueue overflow. Though CrashReporter is enabled (OS X 10.2.3), we see no consistent logs being sent to /Library/Logs/CrashReporter. Clearly the "www" uid suggests either httpd (Apache 1.3.27) or our database (WebDNA 4.5.1) as a possible culprit. But other research has led me to believe that the "IOHIDSystem" mention refers to peripheral devices(USB) or drivers. Please correct my understanding if I'm wrong? Anyone have any suggestions for getting more information from the system to debug this problem? Thanks. Paul Teeter _______________________________________________ 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