9 Nov 06 notes from GCC improvement for Itanium conference call
ON THE CALL: Kenny Zadack (Natural Bridge), Diego Novillo (Red Hat), Vladimir Makarov (Red Hat), Mark Smith (Gelato), Bob Kidd (UIUC), Andrey Belevantsev (RAS), Arutyun Avetisyan (RAS), Mark Davis (Intel), Sebastian Pop (Ecole des Mines de Paris) Agenda: 1) Gelato ICE April GCC track proposed content (Mark S.) 2) GCC4.2 & GCC 4.3 update, alias analysis update (Diego) 3) Scheduler work update, potential new software pipelining project (Andrey) 4) LTO update (Kenny) 5) Superblock work update (Bob) ## 1) Gelato ICE April GCC track proposed content (Mark S.) The content for the GCC track at the upcoming Gelato ICE San Jose (April 15-18) conference was proposed (still need to confirm with several speakers): - ISP-Russian Academy of Sciences: update on scheduler work, discuss progress on new software pipelining work - Martin Michlmayr: compiling Debian using GCC 4.2 and Osprey - Shin-ming Liu: HP GCC and Osprey update - Kenny Zadack - update on LTO - Bob Kidd - update on superblock work - Zdenek Dvorak - update on prefetching work - Diego Novillo - update on alias analysis work - Matthieu Delahaye - update on Gelato GCC build farm - Dan Berlin - GPL2 and GPL3 presentation ## 2) GCC4.2 & GCC 4.3 update, alias analysis update (Diego) GCC 4.2 & 4.3 update: - 4.2 has branched. Likely release in early 2007. Many major pieces of work being scheduled for GCC 4.3 (SSA across the callgraph, overhaul dataflow in the backend, overhaul SSA form for memory, reduce memory footprint in the IL, autoparallelization, new vectorization, new interprocedural optimizations, etc). The full list is at http://gcc.gnu.org/wiki/GCC_4.3_Release_Planning Alias analysis update: -- no changes to analysis, representation of aliasing is being modified for 4.3 ## 3) Scheduler work update, potential new software pipelining project (Andrey) We have merged all major features of selective scheduling and are tuning it for Itanium. We use a set of small benchmarks to analyze the performance of the scheduler. At this moment we are neutral on half of the benchmarks, we get 3% speedup on linpack and 5% speedup on mgrid. We have fixed all of >1% regressions except dhrystone, which regresses on 4% due to alignment issues. Major part of the bugs we've fixed is because bundling and instruction choosing mechanism are tightly coupled with the Haifa scheduler, and we need to support both schedulers at the same time. We plan to proceed with tuning and implement the driver for software pipelining for the next month. We also plan to fix swing modulo scheduling to make it work on ia64 and improve it by propagating data dependency information to RTL. We plan to discuss this project on the GCC mailing list in a few weeks. Comments by Vladimir: - About software pipelining for Itanium: It is completely broken, more accurately it never worked for Itanium. Because it is very (probably most) important optimization for Itanium, after making SP working it should be switched on by default at least for Itanium to keep it working. GCC has a lot of optimizations which are not on by default and they have tendency to be broken sometime. About insn bundling and insn scheduler hooks for Itanium: Usually the scheduler hooks are very few lines. This is not case for Itanium.I believe there is a potential for generating better quality code by improving bundling and hooks. Unfortunately, the code is "spaghetti" code which is hard to understand. ## 4) LTO update (Kenny) The LTO branch has made some progress: there is work underway by Mark Mitchell, Sandra Loosemore, and myself to serialize three codes and the declarations and types into sections inside the .o files. I have stopped working on this temporarily while I complete the work on the dataflow branch. I expect that after the first of the year there will be a lot of progress on this. We should soon be able to serialize code and compile it at link time. This will still leave many problems open for other to help with, including more aggressive optimizations, as well as providing some mechanism to distributing/parallelizing the compilation. ## 5) Superblock work update (Bob) I'm merging mainline into the ia64-improvements branch. As soon as that is finished, I will run a regression on the Superblock pass and prepare a patch to submit to gcc-patches. This work has been on the back burner for the past couple weeks due to an upcoming paper deadline. I'm writing a paper documenting the changes we made to IMPACT's intermediate representation to allow interprocedural analysis to be performed more easily. We were able to extend IMPACT's IR from one stored completely in memory to one that can be stored partially in memory and partially on disk. This allows us to reduce the memory requirements for the compiler when processing modern, large program. This paper will lik
ANNOUNCE: Gelato ICE GCC track, San Jose, CA, April 16-18, 2007
The following GCC track is part of the Gelato ICE (Itanium Conference & Expo) technical program, April 16-18, 2007, San Jose, CA. All interested GCC developers are invited to attend . A working list of speakers and topics can be found here: This year there is a strong focus on Linux. Andrew Morton and Wim Coekaerts, Senior Director for Linux Engineering at Oracle, are keynote speakers. In addition to the GCC track, there are tracks covering the Linux IA-64 kernel, virtualization, tools and tuning, multi-core programming, and research. GCC Track at Gelato ICE: - Update on Scheduler Work & Discussion of New Software Pipelining Work, Arutyun Avetisyan, Russian Academy of Science - GPL2 and GPL3, Dan Berlin, Google - Update on the Gelato GCC Build Farm, Matthieu Delahaye, Gelato Central Operations - Update on Prefetching Work, Zdenek Dvorak, SuSE - Interprocedural Optimization Framework, Jan Hubicka, SuSE - Update on Superblock Work, Bob Kidd, University of Illinois - GCC and Osprey Update, Shin-Ming Liu, HP - Compiling Debian Using GCC 4.2 and Osprey, Martin Michlmayr, Debian - Update on Alias Analysis Work, Diego Novillo, Redhat - Update on LTO, Kenneth Zadeck, NaturalBridge
RE: testing GCC 4.2 on IA64 using Debian as a test suite <--- correction
> Eight IA64 specific and 10 generic GCC defects previously unknown > were identified. All these bugs have been reported to the GCC bug > tracker together with test cases and have all been fixed. Correction/clarification: All IA-64 specific bugs have been fixed.
15 Sept notes from GCC improvement for Itanium conference call
ON THE CALL: Bob Kidd (UIUC), Vladimir Makarov (Red Hat), Mark Smith (Gelato), Wenguang Chen (Tsinghua), Mark Davis (Intel), Diego Novillo (Red Hat), Andrey Belevantsev (RAS), Dan Berlin (dberlin.org), Wen-mei Hwu (UIUC) The call covered: - current status / updates on the 3 improvement areas - brain storming how to help out Dan with the alias analysis improvement TODO list - GCC session at Brazil Gelato meeting - planned presentations (tentative schedule): - Diego Novillo (Monday 10/3, 7:30 to 8pm) - Shin-ming Liu (Monday 10/3, 8 to 8:30pm) - Bob Kidd (Monday 10/3, 8:30 to 9pm - Canqun Yang (Tuesday 10/4, 10 to 10:30am) - general discussion (Tuesday 10/4, 10:30 to 11:30am) - next steps, action items - continue discussion on alias analysis improvements and determine concrete ways to help - review and update IA-64 project list - a discussion on corporate IP needed for GCC -- can companies donate IP to GCC? - planned attendance (from those on the call) - Bob Kidd - Shin-ming Liu - Mark Davis - Wenguang Chen - Diego Novillo - Mark Smith Additional detail can be found below. NEXT MEETING: At Gelato Meeting in Porto Alegre, Brazil, October 2-5, 2005. Bob Kidd: - The patch to move superblock formation is ready to go and tested on x86, x86_64, and ia64. I'll post it to gcc-patches and post a message to the gcc mailing list with details shortly. Dan Berlin: --- The current status of the aliasing work is that work is proceeding on both intraprocedural and interprocedural call clobbering. Most of the infrasturcture to get it to be able to be start work on the actual improved algorithms and use the results is done, or will be done by next week. The main task that it would be helpful to have other people work on would be the "Aliasing Oracle" (pairwise query system for statements) implementation for tree-ssa. The backend already uses a query system exclusively (because it's aliasing is almost exclusively type based). This is an alternate way of representing and accessing the aliasing results that can provide better information (though more expensive to query) than the current representation provides. Our current analyses can actually provide better information that we represent. The eventual plan is to have a hybrid where optimizations use the current "virtual ssa" form to discover what appear to be things that alias, then query the "Aliasing Oracle" to see if it can disambiguate the aliases further if necessary (IE it decides it wants to try to sink a store past that point, or whatever). The implementation can start out simple, (IE even something that simply looked at base + offset of pointers in the two statements, and gave answers, would be a good start) and I'm more than happy to help mentor/introduce anyone to how to work with GCC's tree-ssa form to make this work, if they want to help implement it. Seriously. I'm really more than happy to spend hours teaching someone how to work with this stuff. IBM Research is also more than happy to let me do so. I already help Ken Zadeck a lot with understanding how to get things done in GCC, and he has become quite adept at working with GCC in a short time (< 6 months). If somebody has the resources to provide someone competent to work on aliasing, I'm glad to put them to work for you being productive on GCC :). Vladimir Makarov: - Zdenek Dvorak submitted a patch implementing prefetching on tree-ssa recently. It would be interesting to look how it will work for Itanium. HP and Intel are interesting in gcc performance improvement. If HP and Intel guys know for sure that gcc uses their patents or some their patents could be used for gcc performance improvement, it would be good to give FSF a permission to use the patents in gcc. On the other hand, investigating what patents gcc uses might be a dangerous thing because any compiler most probably uses algorithms patented by other companies. Diego Novillo: -- I spoke briefly about GCC's development process: * Legal issues need to be resolved early. Assign personal and corporate copyright to the FSF. * It's important to participate in the main development lists (gcc and gcc-patches). IRC is a good meeting place, too. * GCC works in three main stages: o Stage 1: Everything goes. o Stage 2: Stabilization. o Stage 3: Bug fixes only. * Work that spans multiple GCC stages is done on CVS branches in the main FSF repository. * I will present at the next Gelato conference. I'll describe GCC's internals, its development process, what is going on and what needs to be done.
RE: 15 Sept notes from GCC improvement for Itanium conference call
Another correction: Shin-ming Liu (HP) attended the meeting. In fact, he led the call :-) > -Original Message- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf > Of Mark K. Smith > Sent: Monday, September 19, 2005 4:00 PM > To: Gelato-GCC; GCC > Subject: 15 Sept notes from GCC improvement for Itanium conference > call > > ON THE CALL: Bob Kidd (UIUC), Vladimir Makarov (Red Hat), Mark Smith > (Gelato), Wenguang Chen (Tsinghua), Mark Davis (Intel), Diego > Novillo > (Red Hat), Andrey Belevantsev (RAS), Dan Berlin (dberlin.org), Wen- > mei > Hwu (UIUC) > > The call covered: > - current status / updates on the 3 improvement areas > - brain storming how to help out Dan with the alias analysis > improvement TODO list > - GCC session at Brazil Gelato meeting > - planned presentations (tentative schedule): >- Diego Novillo (Monday 10/3, 7:30 to 8pm) >- Shin-ming Liu (Monday 10/3, 8 to 8:30pm) >- Bob Kidd (Monday 10/3, 8:30 to 9pm >- Canqun Yang (Tuesday 10/4, 10 to 10:30am) > - general discussion (Tuesday 10/4, 10:30 to 11:30am) >- next steps, action items >- continue discussion on alias analysis improvements > and determine concrete ways to help >- review and update IA-64 project list >- a discussion on corporate IP needed for GCC -- > can companies donate IP to GCC? > - planned attendance (from those on the call) >- Bob Kidd >- Shin-ming Liu >- Mark Davis >- Wenguang Chen >- Diego Novillo >- Mark Smith > > Additional detail can be found below. > > NEXT MEETING: At Gelato Meeting in Porto Alegre, Brazil, October 2-5, > 2005. > > Bob Kidd: > - > The patch to move superblock formation is ready to go and tested on > x86, x86_64, and ia64. I'll post it to gcc-patches and post a > message to the gcc mailing list with details shortly. > > Dan Berlin: > --- > The current status of the aliasing work is that work is proceeding > on > both intraprocedural and interprocedural call clobbering. Most of > the > infrasturcture to get it to be able to be start work on the actual > improved algorithms and use the results is done, or will be done by > next week. > > The main task that it would be helpful to have other people work on > would be the "Aliasing Oracle" (pairwise query system for > statements) > implementation for tree-ssa. The backend already uses a query > system > exclusively (because it's aliasing is almost exclusively type based). > This is an alternate way of representing and accessing the aliasing > results that can provide better information (though more expensive > to > query) than the current representation provides. Our current > analyses > can actually provide better information that we represent. The > eventual plan is to have a hybrid where optimizations use the > current > "virtual ssa" form to discover what appear to be things that alias, > then query the "Aliasing Oracle" to see if it can disambiguate the > aliases further if necessary (IE it decides it wants to try to sink > a > store past that point, or whatever). > > The implementation can start out simple, (IE even something that > simply looked at base + offset of pointers in the two statements, > and > gave answers, would be a good start) and I'm more than happy to > help > mentor/introduce anyone to how to work with GCC's tree-ssa form to > make this work, if they want to help implement it. Seriously. I'm > really more than happy to spend hours teaching someone how to work > with this stuff. IBM Research is also more than happy to let me do > so. I already help Ken Zadeck a lot with understanding how to get > things done in GCC, and he has become quite adept at working with > GCC > in a short time (< 6 months). If somebody has the resources to > provide someone competent to work on aliasing, I'm glad to put them > to > work for you being productive on GCC :). > > Vladimir Makarov: > - > Zdenek Dvorak submitted a patch implementing prefetching on tree-ssa > recently. It would be interesting to look how it will work for > Itanium. > > HP and Intel are interesting in gcc performance improvement. If HP > and Intel guys know for sure that gcc uses their patents or some > their > patents could be used for gcc performance improvement, it would be > good to give FSF a permission to use the patents in gcc. On the > other > hand, investigating what patents gcc uses might be a da
notes, presentations from GCC Improvement for Itanium sessions at Gelato meeting, Oct 3-4, 2005
http://gcc.gelato.org/PortoAlegreMeeting We had excellent sessions and extended time to discuss improving GCC on Itanium. Please refer to the discussion notes for additional information. Many thanks to the presenters and to Shin-ming Liu (HP) for leading the discussion.
10 Nov notes from GCC improvement for Itanium conference call
ON THE CALL: Shin-ming Liu (HP), Vladimir Makarov (Red Hat), Diego Novillo (Red Hat), Mark Smith (Gelato), Andrey Belevantsev (RAS), Arutyun Avetisyan (RAS), Bob Kidd (UIUC), Mark Davis (Intel) The call covered: 1. Setting up GCC branch for Itanium-related work 2. Alias analysis update from RAS and Diego 3. Superblock update from UIUC 4. HP update from Shin 4. Scheduler work from RAS Mark S. will work on securing a Montecito SDV to help test GCC builds. Information about submitting proposals to ISA for GCC work was distributed to the group. Additional call details can be found below. NEXT MEETING: December 8th, 2005. Details will be emailed out prior to the call. Andrey Belevantsev: --- We are working on the first part of the new scheduler infrastructure. This part is a set of routines that gather the instructions available for scheduling. We'd put this on the ia64 branch as soon as this would be ready, hopefully in a couple of weeks. I've tested the tree prefetching pass from the killloop-branch, written by Zdenek Dvorak some time ago. At that time, the branch wasn't stable enough, so SPEC INT results didn't show much. But the pass is a win for SPEC FP tests (8 of 14 tests worked, ~10% mean, from 6% to 40%). I've also tested the killloop branch together with the speculation patch, and the speedups when compared to the pristine branch are more or less the same (as when comparing head+speculation against head) for SPEC FP. The prefetching though helps the speculation for SPEC INT tests (bzip2 speedup is increased by 6%). Probably it's time to redo this testing, because the killloop branch has significantly changed since then. We'll send the alias propagation patch to the list in a few days. The patch will be put on the ia64 branch then, as well as the other ready patches. We've got the confirmation from the FSF for three of us who participated in the last project (Maxim Kuvyrkov, Dmitry Melnik and myself). So to start working on the branch we'll need just the auth tokens in the repository. Bob Kidd: - There's not much new to report on the Superblock scheduling work. I received a cleaned version of the patch from Steven Bosscher, but I haven't had time to look at it yet. I'm working with Diego to set up a branch from the FSF tree and will look into setting up a machine to periodically do performance regressions. Shin-ming Liu: -- HP has done a few things in the past month. We packaged and tested GCC4.0.2 for HP-UX and the bits should be ready for posting shortly. We also start working with GCC community looking into the creation of binary IR files for GCC. It helps the IPO effort in GCC. Vladimir Makarov: - I told about importance of early access to machines based on new Itanium chip (Montecito) and documentation for gcc developers trying to improve gcc for Itanium. As for Mark Davis remark about rewriting RTL optimizations, I told that it can not be done easily. RTL is too complicated. On Andrew Macleod and my etsimation only writing a new good register allocator is at least 2 years project. Significtant simplifying rtl or usage another IR is even more complicated task than introducing Tree-SSA because machine description is very tied to RTL. As an example, combine pass is based on outdated work of Fraser, Proebsting etc. They proposed to combine several intermidiate data dependent insns into one machine insn with possible insn spliting (that what define_split pattern serves in machine desc). Since that work they proposed fast and optimal solution of code selection task with their BURG and IBURG system (finding minimal cost cover of tree expression by machine insn patterns). Moving to this algorithm needs significant simplifying RTL (one rtl insn should be no more than part of machine insn or just one machine insn). That means rewritting all machine desc files (simplier define_insn), removing define_split patterns, rewriting other optimizations (e.g. reload assumes that all moves/stores/loads of one mode is described by one define_insn pattern). Diego Novillo: -- My focus over the last few weeks has been OpenMP and fixing bugs for the 4.1 release. I will create a branch so that folks can put their work in it. I need people to mail me so that I can coordinate write access with them. We discussed briefly some of the activity geared towards improving GCC's backend. The problem is well understood and various relatively independent efforts are moving GCC in the right direction (IPA, dataflow analysis improvements, scheduling, register allocation, move high-level aliasing information into the backend). It will probably take a few releases to fix most of the glaring problems.
8 Dec 05 notes from GCC improvement for Itanium conference call
ON THE CALL: Shin-ming Liu (HP), Vladimir Makarov (Red Hat), Diego Novillo (Red Hat), Mark Smith (Gelato), Bob Kidd (UIUC), Mark Davis (Intel) A fair amount of time was spent discussing the pros and cons of LLVM vs. LTO. Keep in mind that the next Gelato conference is coming up in April 06. If you would like to speak in the GCC track, please contact Mark Smith. Comments from each participant on the call can be found below. NEXT MEETING: January 12th, 2006. Details will be emailed out prior to the call. Bob Kidd: - The ia64-improvements branch is up. I haven't had time to do much with it, but I've checked it out and bootstrapped. I applied Steven Bosscher's revised superblock patch, bootstrapped and tested it. I'm looking at updating the branch to either 4.2 or CVS head and will check in the superblock patch. Shin-ming Liu: -- * HP has posted the GCC 4.0.2 source and binary package on HP site for HP-UX in November * HP is working toward the similar posting for Linux in the near future * HP is actively participating in the LTO activity in GCC community * HP has started the alternative backend effort based on Open64 with several universities Vladimir Makarov: - I asked Bob to tell more details about superblock scheduling he made. Because the current version does not take predication into account, I told that to get better results it is good to combine predication and superblock forming in future. Also superblock scheduling could work without profile information as it does currently according to Bob's review. Gcc has decent evaluation of branch probabilities. It would be reasonable to try how superblock scheduling will work with the branch probabilities evaluation. As for work status of ISP RAS team, they work on infrastructure for another insn scheduling. It is a big project. They sent patch for improvement of aliasing analysis for ia64. Diego Novillo and Daniel Berlin are reviewing the patch. They'll send patch for speculation support for review soon. It is control and data speculation support with recovering code. Now it is necessary only for ia64 but it will be useful for future Intel x86 and x86_64 processors which as I heard will have speculation support too. As for LLVM, I told it is a very interesting research project. In any case it will help to gcc finally and itself. We will see will it be finally adopted in gcc. At least, it have more chances for that than OpenRC with its WHIRL because transition of copyrights to FSF of LLVM is more probable than one of OpenRC from SGI. But adoption of LLVM should not stop other project therefore LTO is a right thing to do. The competition is good and there will be more chance to have intermodule optimizations finally. Additionally to the obstacles to adopt LLVM mentioned by Diego, I named usage of C++ (although it has advantages too) and patents. LLVM should be checked for usage of compiler patents. Gcc people avoided many patents especially from Microsoft. We can not be sure right now about LLVM. Diego Novillo: -- I talked about the new developments in GCC for doing link-time optimizations. We discussed both the LTO proposal and LLVM. Although it's not clear at the moment which proposal will end up being adopted, we all agreed this is an excellent sign that GCC will start evolving in that direction. I also talked about the new ia64-improvements branch and Dmitry's alias patch. The ia64-improvements branch is ready for people to use. I have started reviewing Dmitry's patch and will try to send feedback in the next few days. Mark Smith: --- I will continue to work with Intel to secure a Montecito SDV to use for GCC builds. Itanium Solutions Alliance has funded Bob Kidd's GCC superblock work. I also talked about the upcoming Gelato conference in April 2006. We want to have a strong GCC track at the meeting. Please consider attending. Shin-ming and I will work together on designing the track. If you would like to speak, please contact one of us.
Draft: 12 Jan 06 notes from GCC improvement for Itanium conference call
ON THE CALL: Shin-ming Liu (HP), Vladimir Makarov (Red Hat), Diego Novillo (Red Hat), Mark Smith (Gelato), Bob Kidd (UIUC), Andrey Belevantsev (RAS), Arutyun Avetisyan (RAS) Dan Berlin (IBM) was not able to join the call but did email an aliasing update (see below). We spent some time up front discussing the GCC track at the upcoming Gelato ICE conference (www.gelato.org/meeting). Potential topics and speaker candidates include: 1) Parallel programming with GCC - Diego Novillo 2) Aliasing - Diego Novillo or Dan Berlin (Diego to coordinate with Dan) 3) LLVM - Chris Lattner? (Mark will approach Chris) 4) LTO - Mark Mitchell? (Shin-ming will approach Mark M.) 5) ORC back end for GCC - Shin-Ming Liu 6) Superblock work - Bob Kidd 7) GCC IP issues - Dan Berlin? (Diego will ask Dan) 8) NUDT update - Canqun Yang? (Mark will approach Canqun) 9) RAS update - Andrey Belevantsev and Arutyun Avetisyan 10) Intel micro-architecture talk - Cameron McNairy Also we discussed holding a face-to-face GCC Improvement meeting in mid August in Moscow. Expected benefits: strong participation from Europe GCC community, Intel Moscow, and RAS. Updates from call participants can be found below. NEXT MEETING: Feb 16th, 2006. Details will be emailed out prior to the call. Dan Berlin: --- Update on aliasing: Improvements to the actual points-to analyzer from the improved-aliasing-branch of gcc have been merged into mainline. This includes: 1. An initial implementation of interprocedural pointer analysis (though this is disabled by default for the moment until i finish some follow up patches). 2. Significantly better results for pointers inside structures. 3. Not punting on structures that contain arrays (we used to treat all the fields like a single variable). There is work by Richard Guenther on treating small arrays as if they were structures so we can get better information about them (without resorting to pair based querying). Work on the aliasing query system for those optimizations that want to disambiguate results for specific stores/loads better than we do now will start soon on IAB. Call clobbering changes from IAB are in the process of being merged into the mainline. This mainly consists of much better non-context sensitive intraprocedural call clobbering information, combined with better use of the interprocedural clobber information for static variables that we compute. Once the aliasing query system is done, i will revisit this and implement some context-sensitive analysis for call clobbering to get better results. I also reviewed the patch for propagating points-to sets to RTL, and sent the review to the list. Andrey Belevantsev: --- Dan Berlin has reviewed the alias propagation patch about two weeks ago. Diego mentioned that his concerns about the patch are in line with Dan's. I had a chance to read the review only recently due to the mail problems of our provider. Dan has pointed out to the implementation flaw of the patch, which is using varrays for storing points-to sets instead of bitmaps. This should be fixed to avoid memory and compilation time issues. There are some other less serious issues and the coding style issues. We are going to fix all problems pointed to by Dan and resend the patch. We've sent five parts of the speculation patch to the gcc-patches list, and Vladimir Makarov has taken them for the review. Some issues were already raised in the list, mentioning documentation issues and the effects of the patch for the other platforms. For the x86-64, the influence of the patch is pretty neutral. We're going to address the issues Vlad would probably raise in his review and then resend the patch. We're advancing in the implementation of the new scheduler. We're working on the code motion part, which fixes the scheduling DAG and the data sets after the selected insruction has been scheduled. Part of those is already implemented and under testing. Another part we're starting to work on is the selection of the best instruction for scheduling from the set of available ones. We plan to adopt the corresponding routines from the haifa scheduler. Bob Kidd: - The Superblock formation pass now duplicates the loop header to allow Superblock loops to be unrolled. On a trivial test case, this improves performance on x86_64 and PPC. I'm currently benchmarking the patch on SPEC. Vladimir Makarov: - I am working on review of ISP RAS patch for control and data specualtion support. In general the patch is ok. As for idea of gelato meeting devoted to gcc in Moscow, that is a not bad place because gcc developers from Europe could be there. But probably time (August) is not so good. Diego Novillo: -- I briefly described the OpenMP implementation and what it requires in terms of runtime support. It is now being merged into mainline to be released with GCC 4.2. The Fedora Core GCC 4.1 compiler will also include it. It should work well on Itani
ANNOUNCE: GCC track at Gelato ICE conference
Dear GCC Community, I wanted to let the GCC list know that the Gelato Federation is holding a conference April 24-26 in San Jose. It is called Gelato ICE (Itanium Conference & Expo). http://www.gelato.org/meeting The technical program is very robust (over 50 technical presentations) covering a variety of Itanium-related topics: compilers, architecture, reliability, security, java, virtualization, kernel, memory management, and high-performance libraries to name a few. Of particular interest to this group are the compiler presentations/sessions. Currently there are ten presentations scheduled for the GCC track: * LLVM - Chris Lattner * LTO - Mark Mitchell * Aliasing update - Dan Berlin * Parallel programming with GCC - Diego Novillo * ORC back end for GCC - Shin-Ming Liu * Superblock work - Bob Kidd * Russian Academy of Science scheduler improvement update - Andrey Belevantsev and Arutyun Avetisyan * Intel micro-architecture talk - Cameron McNairy * Spec2006 - Gerolf Hoflehner (provided Spec2006 has been release at conference time) * GCC IP issues - Dan Berlin We also have a round table scheduled with the Intel icc compiler team and a talk on icc changes related to Montecito. Montecito is the next generation of Itanium due out in a few months. Everyone is welcome to attend. Check http://www.gelato.org/meeting for more details. Mark
16 Feb 06 notes from GCC improvement for Itanium conference call
ON THE CALL: Shin-ming Liu (HP), Vladimir Makarov (Red Hat), Diego Novillo (Red Hat), Mark Smith (Gelato), Bob Kidd (UIUC), Andrey Belevantsev (RAS), Arutyun Avetisyan (RAS), Mark Davis (Intel) We spent some time up front discussing the GCC track at the upcoming Gelato ICE conference (www.gelato.org/meeting). Confirmed topics/speakers include: * LLVM - Chris Lattner * LTO - Mark Mitchell * Aliasing update - Dan Berlin * Parallel programming with GCC - Diego Novillo * ORC back end for GCC - Shin-Ming Liu * Superblock work - Bob Kidd * Russian Academy of Science scheduler improvement update - Andrey Belevantsev and Arutyun Avetisyan * Intel micro-architecture talk - Cameron McNairy * Spec2006 - Gerolf Hoflehner (provided Spec2006 has been release at conference time) * GCC IP issues - Dan Berlin Updates from call participants can be found below. NEXT MEETING: March 16th, 2006. Details will be emailed out prior to the call. Andrey Belevantsev: --- We've fixed the alias propagation patch according to Dan Berlin's suggestions and sent it to the list. Namely, bitmaps are now used instead of varrays for storing points-to information; global flags are moved to 'struct function' to be function local and avoid interference with PA passes; coding style issues are fixed. Now we are looking for further feedback from Dan and Diego. Vladimir Makarov has approved one patch from the set of speculation patches. Vlad has asked us to wait with committing until he'll review all the rest. That's better because the separate patches don't allow to use speculation. We've also slightly improved the probability patch originally submitted by Peter Steinmetz and committed it (the thread is at http://gcc.gnu.org/ml/gcc-patches/2006-01/msg02223.html). We've implemented code motion routines of the new scheduler we're working on. The routines perform DAG traversal in search of the original forms of the operation being scheduled. When original operation is found, it is either deleted or changed into a simple copy operation. If the operation should be moved through a join point, a bookkeeping copy should be added on all other paths. This logic is implemented in a slightly different way depending on the type of the operation we are moving (sequential or conditional branch). Now we have focused on adapting the scheduling loop routines and the logic of choosing the best instruction from the set of available ones from the Haifa scheduler. The code motion part will be fully tested only after implementation of these "scheduler frontend" routines. Bob Kidd: - There's not much new on the superblock work. I will check the superblock patch into the ia64-improvements tree later today. After mainline is merged into the tree, I'll run some performance numbers. Vladimir Makarov: - I was brief on the last teleconference. I told that it is good to have some info about SPEC2006 and comparison of Itanium gcc and icc on this benchmark. Some discussion about MMU->address bypass for Montecito. Diego Novillo: -- Mostly working on getting the OpenMP implementation in mainline. So far the C and Fortran front end changes have been incorporated. The C++ changes are being reviewed. It will be released with GCC 4.2. I have also been working on changes to the aliasing infrastructure to avoid some of the penalties we pay for representing alias relations directly in the IL. Other than that, it's been the usual maintenance work.
16 Mar 06 notes from GCC improvement for Itanium conference call
ON THE CALL: Shin-ming Liu (HP), Vladimir Makarov (Red Hat), Mark Smith (Gelato), Bob Kidd (UIUC), Andrey Belevantsev (RAS), Arutyun Avetisyan (RAS), Mark Davis (Intel) Diego Novillo (Red Hat) was unable to join the call, but supplied an update to include in these notes. The GCC track at the upcoming Gelato ICE conference now finalized. Gerolf Hoflehner's talk on SPEC2006 had to be canceled because of a delay in its release. A new addition to the GCC track is Arutyun Avetisyan who will give an RAS work overview and start soliciting input for the August 2006 GCC meeting in Moscow. Confirmed topics/speakers for the Gelato ICE GCC track include: * Russian Academy of Science work overview and plans for August GCC meeting in Moscow - Arutyun Avetisyan * GCC IP issues - Dan Berlin * LLVM - Chris Lattner * LTO - Mark Mitchell * ORC back end for GCC - Shin-Ming Liu * Aliasing update - Dan Berlin * Russian Academy of Science scheduler improvement update - Andrey Belevantsev * Superblock work - Bob Kidd * Parallel programming with GCC - Diego Novillo * Intel micro-architecture talk - Cameron McNairy For a detailed list of confirmed speakers and topics for Gelato ICE 2006, visit: www.gelato.org/meeting#agenda Updates from call participants can be found below. NEXT MEETING: At the Gelato ICE meeting in San Jose, CA, April 24-26, 2006. Andrey Belevantsev: --- Testing the aliasing patch with the latest mainline has revealed the changes in structure aliasing, so we had to rewrite some code that handles variables with structure field tags (SFTs). Now small arrays could also be decomposed onto elements for the sake of better aliasing. The other thing we fixed is more accurate propagation of original tree expressions saved with MEMs during expand. We have sent an updated patch to the gcc-patches list. Vladimir Makarov has approved the speculation patch and provided commetns on the ia64 part of the patch. We have fixed all issues pointed to by Vladimir. After additional regtesting on ia64 and i686, the patch was committed to trunk as rev. 112129. Earlier version of the patch was also bootstrapped and regtested on sparc-solaris. Using the patch on other platforms revealed some bugs (PR26275 and PR26734). The fixes for those PRs are submitted to the list. We have tested the basic features of code motion during this month. To accomplish this task, the main scheduling loop was written. A single iteration of the scheduling loop tries to form a group of instructions, which could be executed in parallel during one cycle (more or less corresponds to the instruction group of IA-64). At first, code motion of entire instructions inside a basic block was tested. Now we are testing interblock motions, which imply possible creation of bookkeeping code. Code motion of conditional branches is now disabled. Our next plans would be enable the code motion of right-hand sides of expressions. The last but not least, our paper proposal for GCC Summit 2006 has been accepted. The paper will talk about new scheduler work, proposed design and current state of implementation. Bob Kidd: - (Bob had his paper proposal for the GCC Summit 2006 accepted. The paper will cover the GCC superblock work in detail.) I checked the Superblock patch into the ia64-improvements tree. This patch has no significant effect on the overall estimated SPEC score for ia64 or ppc, and a slight degradation on x86_64. On IA64, some benchmarks run faster while others slow down. The overall score varies by one point. I'm looking into the changed benchmarks to see what causes the speedup or slowdown. I investigated 300.twolf, which slows down when superblocks are formed at the Tree-SSA level. One function (new_dbox_a) is significantly slower with the superblock patch than without. This function takes a pointer to an integer as an argument and updates the value of that integer inside a hot loop. The loop is structured along these lines: for (hot) if (cond) (biased) a = ... else a = ... *arg += a ... Tail duplication generates two copies of the *arg += ... line, which generates two copies of the load and store of arg. When tail duplication is not done, PRE can move the load and store of arg out of the loop, but it is unable to do this in the superblock loop. My suspicion is that superblock formation needs to fix up the alias info so that later optimizers realize these two loads are the same. Shin-ming Liu - - HP has posted the GCC 4.1 release binary in HP portal for HP-UX: www.hp.com/go/gcc - HP submitted 11 patches to stock gcc and 3 patches to binutil - The Alternative backend project has made reasonable progress. The front end for this compiler still at 3.3.2. Both C and Fortran are functional and achieved the similar performance as ORC 2.1. The current focus is to update the backend to support Itanium C++ ABI. Vladimir Makarov: - Probably Robert Kidd's superblock scheduling in gcc
ANNOUNCE: GCC track at Gelato ICE finalized - April 24 & 25
The following GCC track is part of the Gelato ICE (Itanium Conference & Expo) technical program. All interested GCC developers are invited to attend . A complete list of speakers and topics can be found here: Monday Afternoon (4/24/06): * 2:15 pm - Arutyun Avetisyan The ISP RAS Effort to Improve GCC for Itanium August '06 GCC meeting in Moscow * 3:00 pm - Dan Berlin GCC IP issues * 4:15 pm - Shin-Ming Liu An ORC Backend for GCC * 5:00 pm - Cameron McNairy Itanium 2 and Montecito Microarchitecture Monday Night (4/24/06): * 7:00 pm - Dan Berlin Aliasing in GCC * 7:30 pm - Bob Kidd Superblock Update * 8:00 pm - Andrey Belevantsev An Interblock VLIW Targeted Instruction Scheduler for GCC * 8:30 pm - Diego Novillo Parallel programming with GCC Tuesday Afternoon (4/25/06): * 2:30 pm - Mark Mitchell LTO: A Brief Introduction * 3:30 pm - Chris Lattner LLVM: A Brief Introduction
ANNOUNCE: Gelato ICE GCC presentations available
The following presentations from the Gelato ICE GCC track (April 24 and 25) can be found at: http://www.gelato.org/community/gelato_meeting.php?id=ICE06aprT (see bottom of left index column entitled "Focus on GCC" for links) - The ISP RAS Effort to Improve GCC for Itanium, Arutyun Avetisyan - GCC IP Issues, Dan Berlin - Open64: An Alternative Backend for GCC, Shin-Ming Liu - Aliasing in GCC, Dan Berlin - Superblock Update, Robert Kidd - An Interblock VLIW-Targeted Instruction Scheduler for GCC, Andrey Belevantsev - Parallel Programming with GCC, Diego Novillo - LTO: A Brief Introduction, Mark Mitchell - LLVM: A Brief Introduction, Chris Lattner
ANNOUNCE: Gelato GCC Improvement on Itanium Workshop Summary, 7-8 August, Moscow Russia
A meeting of the Gelato GCC Improvement on Itanium Workgroup took place August 7 & 8, 2006 in Moscow, Russia. The workshop was hosted by the Institute for System Programming at the Russian Academy of Sciences and was sponsored by Intel and HP. http://gcc.gelato.org/MoscowMeeting Compiler experts from the GCC open-source community, Red Hat, SuSE, Intel, HP, and the Gelato Member community discussed specific GCC improvements for the Itanium platform. Several key areas were identified to improve Itanium GCC performance. http://gcc.gelato.org/MeetingNotes Presentations from the workshop (all presentations can be found at http://gcc.gelato.org/MoscowMeeting): * An Architectural Overview of GCC, Diego Novillo * GCC RTL Backend, Steven Bosscher * The IA64 Backend in GCC, Jim Wilson * LTO in GCC, Ken Zadeck * LNO in GCC, Sebastian Pop * PDO in GCC, Jan Hubicka * Itanium Architecture and ICC Tutorial, Mark Davis * Performance Metrics & Measurement, Shin-Ming Liu * CERN Loops Dissected, Sverre Jarp * Alias Analysis in GCC, Diego Novillo * Software Pipelining in GCC, Vladimir Makarov * Data Prefetching in GCC, Zdenek Dvorak * Instruction Scheduling in GCC, Vladimir Makarov * Superblock Update, Robert Kidd * ISP RAS Scheduling Update, Andrey Belevantsev
