Difference between revisions of "User:Kr"
From Apache OpenOffice Wiki
m (Extract building OOo into its own page) |
m (Extracted too much ;-)) |
||
| Line 30: | Line 30: | ||
* http://en.wikipedia.org/wiki/Mac_OS_X | * http://en.wikipedia.org/wiki/Mac_OS_X | ||
| − | == | + | ==Misc== |
| − | [[/Building OOo]] | + | * My Linux Desktop Architects Meeting II presentation slides: [[Media:DAM2-OOo_1_1.odp]] |
| + | * My OOo Conf 2006 presentation slides: [[Media:OOoConf_2006_Cleaning_Up_Multi-Threading.odp]] | ||
| + | * A short presentation prepared for the OOo BOF at EuroOSCON2006: [[Media:EuroOSCON2006-OOo_Architecture.odp]] | ||
| + | * Broken __attribute__((constructor)) for gcc 3.3 | ||
| + | ** http://www.opensource.apple.com/darwinsource/Current/libstdcxx-11/libstdcxx/NEWS | ||
| + | ** http://www.mail-archive.com/debian-gcc@lists.debian.org/msg06180.html | ||
| + | * "exit" libc API call is racy wrt to pthreads on some Systems, see [[User:Kr/exit_race.c]] | ||
| + | |||
| + | ==Notes== | ||
| + | * "pthread_exit" and its relationship to process termination ... | ||
| + | ** "atexit" actually gets called by the last thread in a process | ||
| + | ** "pthread_exit" in "main" would actually be perfect, if it was allowing to pass another return value than zero ... | ||
| + | ** [[/Building OOo]] | ||
==My Efforts== | ==My Efforts== | ||
Revision as of 07:43, 5 September 2007
Contents
Contact
- Mail: mailto:Kay.Ramme@sun.com
Background
- Full name is Kay Ramme
- Uno (UDK) Project Lead of OOo
- Located in Hamburg, Germany
- Employed by Sun Microsystems
Some Thoughts
- User_talk:Kr/Parameterized Headers
- User_talk:Kr/UNO References
- User:Kr/Software Continuum
- User:Kr/On Modularization
- User:Kr/On_Streams
GullFOSS
- My Visions
- ODF being a Disruptive Technology
- Memory Savings by String internalization
- AAWAX or How to Check Out OpenOffice.org fast
- Updated Uno Web Pages
- The Java platform becoming Open Source
Some Scripts
Platforms
Misc
- My Linux Desktop Architects Meeting II presentation slides: Media:DAM2-OOo_1_1.odp
- My OOo Conf 2006 presentation slides: Media:OOoConf_2006_Cleaning_Up_Multi-Threading.odp
- A short presentation prepared for the OOo BOF at EuroOSCON2006: Media:EuroOSCON2006-OOo_Architecture.odp
- Broken __attribute__((constructor)) for gcc 3.3
- "exit" libc API call is racy wrt to pthreads on some Systems, see User:Kr/exit_race.c
Notes
- "pthread_exit" and its relationship to process termination ...
- "atexit" actually gets called by the last thread in a process
- "pthread_exit" in "main" would actually be perfect, if it was allowing to pass another return value than zero ...
- /Building OOo
My Efforts
Effort/Revise OOo Multi-Threading
- Tasks
- review documentation, place tests and testers
- rename Environment.pReserved -> pEnterable
- do threading tests - with component, with the office
- convert rest of office services to be thread unsafe - in progress
- convert dtrans to thread affine?
- Rename Threading Model impl to Threading Framework?
- See if I can optimize the "identity mapping" somehow.
- Possible order of integration
- in progress : VCL thread-transparency 1 - vclthreadtransparency1 - depends: bunoexttm
- fix thread-affinity
- DDE
- Drag&Drop
- Clipboard
- open : Binary Uno new threading architecture - bunonewta - depends: bunoexttm
- in progress : async dialogs 2 - asyncdialogs2 - depends: asyncdialogs
- open : VCL thread-transparency 2 - vclthreadtransparency2 - depends: asyncdialogs2, vclthreadtransparency1
- remove the Solar Mutex
- open : new OOo threading architecture - ooonewta - depends: bunoexttm, vclthreadtransparency2
- in progress : VCL thread-transparency 1 - vclthreadtransparency1 - depends: bunoexttm
Move Uno stuff into the Uno Wiki. See Uno/Effort/Migrate Documentation
- Update udk.openoffice.org to reflect latest wiki changes.
Do a source code inventory
Some Links
Event Driven Programming
- Some comments regarding Linux kevents etc.: http://lwn.net/Articles/225714/ Note: In my understanding the whole discussion is senseless, the way to go for awaiting events should certainly not be fd related, but should just use something as "sigwait" or similar. Please see also Architecture/Proposal/Advanced_Threading-Architecture.
Threading
- OOo Threading-Architecture specification
- Architecture/Proposal/Advanced Threading-Architecture
- COM - Processes, Threads and Apartments - "The key is to use multiple threads only if you are very sure of what you are doing." - http://windowssdk.msdn.microsoft.com/en-us/library/ms693344.aspx
- Designing and Building Parallel Programs - http://www-unix.mcs.anl.gov/dbpp/
- The Future of Programming - http://alarmingdevelopment.org/index.php?p=6
- Programming in the Age of Concurrency: Software Transactional Memory - http://channel9.msdn.com/showpost.aspx?postid=231495
- The Problem with Threads - http://www.computer.org/portal/site/computer/menuitem.5d61c1d591162e4b0ef1bd108bcd45f3/index.jsp?&pName=computer_level1_article&TheCat=1005&path=computer/homepage/0506&file=cover.xml&xsl=article.xsl
- Threads and memory model for C++ - http://www.hpl.hp.com/personal/Hans_Boehm/c++mm/
- Why Threads Are A Bad Idea (for most purposes) - http://home.pacbell.net/ouster/threads.pdf
- Swing: How to Use Threads - http://java.sun.com/docs/books/tutorial/uiswing/misc/threads.html
- A comment by Graham Hamilton on thread-safe toolkits - http://weblogs.java.net/blog/kgh/archive/2004/10/multithreaded_t.html
- An article on the future of CPUs, it comments on multi-threaded applications: http://www.informit.com/articles/article.asp?p=663085&seqNum=6&rl=1
- An article showing ways of utilizing SMP on UNIX: http://www.informit.com/articles/article.asp?p=679545&rl=1
- Futexes are Tricky - http://people.redhat.com/drepper/futex.pdf
