GCC Gathering 2011
Who is who? Can someone fill in the names from top left to lower right?
(name or color of shirt)
- standing
- Matthias Klose, Jeremy Bennett, Ramana Radhakrishnan, unknown, Cameron Palmer, blue, Andrew Haley, Jan Kratochvil, Cary Coutant, Diego Novillo, Phil Muldoon, Zbignew Chamski, Ian Taylor, Dodji Seketeli, Richard Guenther, Joern Rennecke, Petr Machata, Albert Cohen, Bhaskar Janakiraman
- kneeling
- Artem Shinkarov, red, Michael Matz, Doug Kwan, Tobias Burnus, Martin Jambor, Mark Wielaard
- photographing
- Lawrence Crowl
Meeting Summary and Notes
At the start of the meeting, all the attendees voted on the list of topics to discuss and we discussed them in order of popularity. For the least popular topics, we split into four different streams.
The major topics we discussed as a group included:
- Patch review
- GDB
- C++0x vs Optimizers
- Plugin API
GCC migration to C++ (GCC_in_C++.html)
- Community issues
- LTO scalability
- Pre-parsed headers
- Bitfield lowering
Supporting BISC machines (bisc.pdf)
gengtype and C++ (notes available at http://gcc.gnu.org/wiki/Memory_management)
For a detailed account of the discussion and conclusions, read the attached notes (GCCGathering2011.pdf).
Meeting Organization
This page describes scheduling and organizational details for the GCC Developers Gathering at the Google office in London, UK for the weekend of 17-Jun-2011. We will update it with notes in the near future.
We will meet Friday evening, all day Saturday, and Sunday until some time in the afternoon.
The idea is to simply get together and discuss current/future work, coordinate efforts, and perhaps do some collective GCC hacking.
The format is going to be an informal unconference:
- No papers.
- No prepared presentations (unless it's really interesting).
- No attendance fees.
If you are interested in attending, please contact Diego Novillo or Ian Taylor.
You may also subscribe to the mailing list gcc-gathering@googlegroups.com, visit https://groups.google.com/group/gcc-gathering to join.
Meeting details
We will meet at the Google London office from Fri 17/Jun at 18:30 to Sun 19/Jun at 15:00.
LOCATION
- Google's London Office
- Belgrave House
- 76 Buckingham Palace Road
- London SW1W 9TQ
- United Kingdom
SCHEDULE
Friday, June 17
- 6:30 PM - Registration and casual Dinner is served.
- 8:00 PM - Midnight - sessions and conversation.
Saturday, June 18
- 8:00 - 9:00 AM - Breakfast served.
- 9:30 AM - 12:30 PM - Sessions.
- 12:30 - 2:00 PM - Lunch served.
- 2:00 PM - 6:00 PM - Sessions.
- 6:00 - 7:30 PM - Dinner served.
- 8:00 PM- Midnight - sessions and conversation
Sunday, June 19
- 8:00 - 9:00 AM - Breakfast served.
- 9:30 AM - 12:30 PM - Sessions.
- 12:30 - 1:30 PM - Lunch served.
- 1:45 PM - 3:00 PM - Closing session.
We have reserved 4 conference rooms in the office. 3 rooms with 12-18 person capacity and the main room that can hold up to 60 people. Every room is fitted with projectors and there is wireless access, so if you are thinking of projecting a few slides, feel free to bring your laptop.
Topics to discuss
The following table describes some of the proposed discussion topics.
Topic |
Owner |
Description |
Pre-parsed headers |
dnovillo |
Status of pph branch, issues |
Bitfield lowering |
richi |
Status and issues |
sizetypes and HOST-WIDE-INTs |
richi |
Fun (or not) |
Plugin API |
iant |
Let's try to get a plan of record |
Patch review |
iant |
Always worth discussing |
Community |
iant |
General leadership, including new contributors |
C++(0x) vs the optimizers |
pcarlini |
vector<int*> vs vector<unsigned int*> vs LTO; 'final', status of devirtualization? |
LTO scalability for large apps |
honza |
What problems we need to solve here |
Cgraph/Varpool and symbol table |
honza |
Fun |
Too many active release branches |
richi |
Definitely Fun |
Community |
iant |
How do we make project-wide technical decisions? |
libjava |
mjw |
can we reduce its size for bootstrap testing? |
java |
dnovillo |
should we remove it from primary languages? |
IPA-CP |
martinj |
Short overview of my new IPA-CP, what more can be achieved with it and what has to be done to get there. |
GDB |
pmuldoon |
Interoperability/issues/open floor/improvements |
C++ & gengtype |
bstarynk |
C++, GTY, gengtype (Basile Starynkevitch) |
C++ style and migration |
crowl |
writing and using C++ in gcc |
BISC architecture support |
zchamski |
proper backend support for exposed-pipeline architectures (DSP, VLIW, delay slots) |
Decl state tables vs streamer cache. Unify? |
dnovillo |
Do we really need two referencing mechanisms? Both seem to solve the same problem |
Hardened code generation |
bjanakiraman |
Dealing with performance implications of hardened code gen for mobile OS (Chrome OS, Android) |
tkoenig |
All things Fortran |