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

Re: mapping user-space memory into kernel, reading, crashing
by Andrew Gallatin 04 Apr '03

04 Apr '03

Paul R.Potts writes: > userland_mem_desc = IOMemoryDescriptor::withAddress ( > user_space_pointer, > (unsigned long) user_space_size, > kIODirectionIn); <..> > > However, it appears that an arbitrary point, somewhere between 700 > words and 1000 words, the kernel will always crash. (I have not used > trial-and-error to narrow it down to the exact count; reading up to 700 > or 2800 bytes into the block succeeds; reading 1000 words fails). Mac's have 4096 byte pages, so it sounds like you're crashing when you cross a page boundary. What is "user_space_size" for your experiments? Have you verified that the kernel is not attempting to access data "user_space_size" bytes past the beginning of the segment? Drew _______________________________________________ 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

mapping user-space memory into kernel, reading, crashing
by Paul R.Potts 04 Apr '03

04 Apr '03

Hi all, I have code (based on Apple's samples) that maps a chunk of user-space memory into the kernel and attempts to read from it. The user-space memory is a local array defined like: UInt8 buf[400000] (400,000 bytes) I'm passing this by pointer into the kernel using an IOConnectMethodScalarIScalarO call; the kernel maps it like so: // construct a memory descriptor for the client memory address userland_mem_desc = IOMemoryDescriptor::withAddress ( user_space_pointer, (unsigned long) user_space_size, kIODirectionIn); if (NULL == userland_mem_desc) { IOLog("IOMemoryDescriptor::withAddress returned NULL\n"); goto finish; } // wire it and make sure we can read it err = userland_mem_desc->prepare (kIODirectionIn); if (kIOReturnSuccess != err) { IOLog("userland_mem_desc->prepare failed(%lx)\n", (unsigned long) err); goto finish; } // map() will create a mapping in the kernel address space. userland_mem_map = userland_mem_desc->map(); if (NULL == userland_mem_map) { IOLog("userland_mem_desc->map() failed(%lx)\n", (unsigned long) err); goto finish; } // get a usable virtual address for this chunk of memory kernel_mapped_userland_memory_addr = (void *) userland_mem_map->getVirtualAddress(); if (NULL == kernel_mapped_userland_memory_addr) { IOLog("userland_mem_map->getVirtualAddress returned NULL!\n"); goto finish; } This all seems to work great. The kernel can then read the expected data out of the chunk of mapped memory. I do this as words (unsigned longs). However, it appears that an arbitrary point, somewhere between 700 words and 1000 words, the kernel will always crash. (I have not used trial-and-error to narrow it down to the exact count; reading up to 700 or 2800 bytes into the block succeeds; reading 1000 words fails). Is there some arbitrary limit on the size of the chunk of user-space memory I can successfully map with the above call? My impression was that the mapping and preparation calls would do what was necessary to map all the required pages, make them readable, etc. I have a workaround; the workaround is to allocate the memory in the kernel and map it into user space, then call the kernel and tell it to read the data back out. This is working, so the question is not urgent, but I am wondering if I'm missing some information. Thanks, Paul _______________________________________________ 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: string handling broken in dev_finddir(), bsd/miscfs/devfs/devfs_tree.c
by Daniel Mack 04 Apr '03

04 Apr '03

On Mon, Mar 17, 2003 at 11:07:51PM +0000, Quinn wrote: > >There is a bug in the string handling in dev_finddir(), > >bsd/miscfs/devfs/devfs_tree.c > > Please file a bug report so that the relevant engineering folks get > to see this. > > <http://developer.apple.com/bugreporter/> I did that right after your suggestion, without any response yet. My bug still has "Open/Analyze" state (filed as #3200736). Is there anybody actually working on the Kernel at Apple? Another thing I wonder about is where I can get the current Kernel version (6.4) in source code. All the repositories I found so far only have 6.3 in their main branch. Do I really have to suffer all the bugs others fixed in the last months when making own improvements in this code? Or did I miss anything? Daniel _______________________________________________ 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: Attaching to /dev/kmem with gdb
by Garth Cummings 04 Apr '03

04 Apr '03

Hi Brian, On Thursday, April 3, 2003, at 09:42 PM, Brian Tabone wrote: Now I know you can't attach directly to a running kernel as you would a program, since halting the kernel would halt you. You can have the running kernel stop to allow an attach from a remote GDB session, then resume the kernel and debug it much as you would an application. Here are a couple of documents you can read that explain this: HelloDebugger <http://www.opensource.apple.com/projects/documentation/howto/html/ kext_tutorials/hello_debugger.html> Inside Mac OS X: Kernel Programming <http://developer.apple.com/techpubs/macosx/Darwin/General/ KernelProgramming/build/chapter_18_section_7.html> Good luck, --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: Can not find TCPLogger example sources
by Justin Walker 04 Apr '03

04 Apr '03

On Friday, Apr 4, 2003, at 03:01 US/Pacific, Sebastien Gaide wrote: Hello, I'm looking for the sources of the TCPLogger example, I have searched the cvs web site with no success. I have searched the archive of this list, and find this url : http://www.opensource.apple.com/cgi-bin/registered/cvs/src/live/ TCPLogger but it seems to be outdated... What does 'outdated' mean? 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

Re: Attaching to /dev/kmem with gdb
by Andrew Gallatin 04 Apr '03

04 Apr '03

Sam Vaughan writes: > Hi Brian, > > Do you have to connect to the kernel on the same machine? It's much > easier to mess around inside a kernel if you do it from a remote > machine. There are detailed instructions for how to do this towards Its much easier to just use the same machine, especially if you have only one. ;) Drew _______________________________________________ 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: Attaching to /dev/kmem with gdb
by Andrew Gallatin 04 Apr '03

04 Apr '03

Paul Ripke writes: > > gdb -s /mach.sym /dev/kmem > > > > Now I know you can't attach directly to a running kernel as you would > > a program, since halting the kernel would halt you. But, from what I > > understand you can attach to the running image to examine the memory, > > without halting that image. Perhaps I am mistaken? > > I'm led to believe that this is not supported with the Mac OS X gdb. It > would be nice - I find it handy on NetBSD, AIX and Tru-64. > And FreeBSD, and Solaris, and.. well.. pretty much every other unix under the sun. In OS-X's favor, it probably has the best remote debugging facility around. You can do what you want by breaking into the debugger and attaching a remote gdb over the network. And when you do this, you get a consistant view of the world, because the kernel is halted. My OS-X debugging wishlist would be: - symbolic stack traces on kernel panics. This requires a little magic from the kernel linker, but hey, if FreeBSD, Solaris, AIX, etc can do it, OS-X should be able to do it too. It would be really nice, if a customer reported a panic, to be able to see where it happened. - local attachment via gdb -k to /dev/kmem. - crashdumps. Since its open-source, I suppose people are free to implement these things... Drew _______________________________________________ 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: Attaching to /dev/kmem with gdb
by Sam Vaughan 04 Apr '03

04 Apr '03

Hi Brian, Do you have to connect to the kernel on the same machine? It's much easier to mess around inside a kernel if you do it from a remote machine. There are detailed instructions for how to do this towards the end of Apple's Kernel Programming document: <http://developer.apple.com/techpubs/macosx/Darwin/General/ KernelProgramming/build/index.html> Sam On Friday, Apr 4, 2003, at 07:00 Europe/London, darwin-kernel-request(a)lists.apple.com wrote: Now I know you can't attach directly to a running kernel as you would a program, since halting the kernel would halt you. But, from what I understand you can attach to the running image to examine the memory, without halting that image. Perhaps I am mistaken? _______________________________________________ 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: Attaching to /dev/kmem with gdb
by Paul Ripke 04 Apr '03

04 Apr '03

On Friday, Apr 4, 2003, at 15:42 Australia/Sydney, Brian Tabone wrote: List, I have heard of old school UNIX hacks attaching to live kernels with gdb, allowing them to see values of variables and changing those values on the fly. I am curious as to how one could do this on Darwin. I attempted to attach to /dev/kmem via gdb, as root. All I would get back was a complaint that /dev/kmem could not be found (I looked and its there all-right). What I can gather is that /dev/kmem is a character special file, and is not visible to gdb in a direct fashion. I used the following gdb -s /mach.sym /dev/kmem Now I know you can't attach directly to a running kernel as you would a program, since halting the kernel would halt you. But, from what I understand you can attach to the running image to examine the memory, without halting that image. Perhaps I am mistaken? I'm led to believe that this is not supported with the Mac OS X gdb. It would be nice - I find it handy on NetBSD, AIX and Tru-64. On NetBSD: ksh$ gdb GNU gdb 5.0nb1 Copyright 2000 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i386--netbsdelf". (gdb) file /netbsd Reading symbols from /netbsd...(no debugging symbols found)...done. (gdb) symbol-file /netbsd.gdb Reading symbols from /netbsd.gdb...done. (gdb) target kcore /dev/mem #0 0xc03f7700 in lwp0 () On Darwin: ksh$ gdb GNU gdb 5.3-20021014 (Apple version gdb-250) (Sat Dec 7 02:14:27 GMT 2002) Copyright 2002 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "powerpc-apple-macos10". (gdb) file /mach_kernel Reading symbols from /mach_kernel...done. (gdb) symbol-file /mach.sym Reading symbols from /mach.sym...done. (gdb) target kcore /dev/mem Undefined target command: "kcore /dev/mem". Try "help target". Cheers, -- Paul Ripke Unix/OpenVMS/TSM/DBA 101 reasons why you can't find your Sysadmin: 68: It's 9AM. He/She is not working that late. -- Koos van den Hout _______________________________________________ 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

Can not find TCPLogger example sources
by Sebastien Gaide 04 Apr '03

04 Apr '03

Hello, I'm looking for the sources of the TCPLogger example, I have searched the cvs web site with no success. I have searched the archive of this list, and find this url : http://www.opensource.apple.com/cgi-bin/registered/cvs/src/live/ TCPLogger but it seems to be outdated... can someone help me to find these files? Thank you very much. Sebastien. _______________________________________________ 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