- linaro-toolchain - lists.linaro.org

Re: [RFC] NEON vs. ARM register selection

by Steven Bosscher

> The basic idea is that we add a new RTL optimization pass (or two) that > assesses the usage of pseudo registers, and makes recommendations about > what register class each should end up in, if there's a choice. These > recommendations would then be used by later passes to get a better use > of NEON. I might call this the "prealloc" pass, or something. That sounds very much like the pre-reload that "new-ra" had at one point (http://gcc.gnu.org/viewcvs/branches/new-regalloc-branch/gcc/pre-reload.c). The problem with pre-reload for new-ra was that it was basically reload instead of something nicer and cleaner. It also only ran just before the register allocator, which is too late for the problem you are trying to solve. > Firstly, for each pseudo-register in a function, the pass would look at > the insn constraints for each "def" and "use", and see how the registers > relate to one another. This might determine things like "if rN is in > class A, then rM must be also in class A". At SUSE I tried to do this with the webizer pass (web.c). I wrote down the ideas we implemented at the time (see http://gcc.gnu.org/ml/gcc/2005-01/msg00179.html): - web class, to replace regclass and choose register classes webs instead of pseudos. This also includes splitting webs if a register in a web really wants to be in two different classes to satisfy constraints in two different insns. Right now, as far as I understand, regclass just picks one and lets reload figure out how to fix up that mistake. - A semi-strict RTL mode. Right now there is just strict and non-strict. On the branch there is a semi-strict mode which is the same as strict RTL except that pseudo-registers are still allowed. - pre-reload (which is related to web class) to make sure as many insn constraints as possible are satisfied before the register allocator goes to work. Basically, after pre-reload the insns stream should be in semi-strict RTL form. I used the webizer to unify defs and uses. I would split a web if it needed multiple register classes (I inserted a mov, without checking that a move existed from the source to the target register class), and I put pseudos r1 and r2 in the same register class if there was an insn (set (r1) (r2)) somewhere. The selection of the register classes had a cost function, but I used rtx_cost, which is not very effective, really. But I never took this experiment very far because for x86-64 the plan didn't work as well as I had hoped. I don't remember the details, but the biggest problem I had with the experimental implementation of these ideas (apart from lots of trouble with recog for semi-strict RTL) was that there is a bit of an ordering problem between combine on the one hand, and web-based register classes. If you assign classes too early and don't allow things to change, then combine fails too often. If you assign register classes after combine, you may not get the instructions selected the way you want them to be. This was when GCC still had the old local-alloc.c and global.c allocators. Things may be different (better) with IRA and the upcoming LRA stuff. If you plan to work on this, I would suggest you discuss the plan on the GCC mailing list also, with Jeff Law and Vladimir Makarov in CC because they are working on a reload rewrite (LRA). Ciao! Steven

13 years, 5 months

1
0
0 0

[ACTIVITY] weekly status

by Ken Werner

Hi, OpenEmbedded: * Worked on the meta-linaro layer and added libgcc and crosssdk recipes to satisfy some bitbake dependencies * I had to apply a few patches to build the linaro toolchain the OE way (mostly gcc configury) * successfully built the sato and Qt images * Moved on to test the February release of the linaro binary toolchain and (probably) and hit an issue with unaligned SD card images to used with QEMU * the guest kernel fails with: attempt to access beyond end of device * /proc/partitions shows different block sizes (host vs. guest) * the image size gets calculated on the fly by OE * patch posted that introduces allows to specify a rootfs size alignment * not seen on trunk as they use IDE * Started to rebase the linaro-meta layer against current OE-core * created https://wiki.linaro.org/KenWerner/Sandbox/OEMetaLinaroCard based on the existent card of David R. Regards, Ken

13 years, 5 months

1
0
0 0

[ACTIVITY] Feb 27 - Mar 2

by Ulrich Weigand

== GCC == * Fixed mainline regression causing ICE in certain outer-loop vectorization cases. * Merged fwprop-subreg patch into Linaro GCC 4.7. * Completed patch to generate usat/ssat instructions where appropriate; checked into GCC mainline. Merge requests to Linaro GCC 4.6 and 4.7 pending. * Ongoing work on improving end-of-loop value computation. Mit freundlichen Gruessen / Best Regards Ulrich Weigand -- Dr. Ulrich Weigand | Phone: +49-7031/16-3727 STSM, GNU compiler and toolchain for Linux on System z and Cell/B.E. IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter | Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen | Registergericht: Amtsgericht Stuttgart, HRB 243294

13 years, 5 months

1
0
0 0

[Activity ] Week 9

by Ramana Radhakrishnan

==Progress=== * Finished off PGO patch - sent upstream. * Finished off the ABI tests - sent upstream. * Investigated fixes for LP 942307 - a problem with kernel builds for android. Backported a fix from Uli last year. * Upstream patch review. * Small configury done for SPEC2k as far as HC partitioning goes. * Some Android benchmark investigations. * Recovered from a broken upgrade on my laptop from natty to oneiric on my laptop and then went all the way to Precise. It works reasonably ! === Plans === * Commit all approved and tested patches. * Check on hc partitioning results from SPEC2k and make sure there is an improvement and the feature works ! * Investigate https://bugs.launchpad.net/gcc-linaro/+bug/924726 in a little more detail. * Get back to partial-partial PRE. Absences. * 1 week holiday sometime before that - to be booked. * Linaro Connect Q2.12 - May 28 - June 1 - travel booked - hotel to be booked.

13 years, 5 months

1
0
0 0

Agenda for next performance call

by Ramana Radhakrishnan

is now here https://wiki.linaro.org/WorkingGroups/ToolChain/Meetings/2012-03-06 Please add any topics that you might consider interesting for next time. Ramana

13 years, 5 months

1
0
0 0

[ACTIVITY] report week 09

by Peter Maydell

Current Milestones: || || Planned || Estimate || Actual || ||cp15-rework || 2012-01-06 || 2012-??-?? || || (new blueprints & reestimate for this one pending) Historical Milestones: ||a15-usermode-support || 2011-11-10 || 2011-11-10 || 2011-10-27 || ||upstream-omap3-cleanup || 2011-11-10 || 2011-12-15 || 2011-12-12 || ||initial-a15-system-model || 2012-01-27 || 2012-01-27 || 2012-01-17 || ||qemu-kvm-getting-started || 2012-03-04?|| 2012-03-04?|| 2012-02-01 || == cp15-rework == * ploughing through conversion of cp15 registers to new design: patchset now 20 patches long, still TODO crn={0,1,6,7,9} == other == * reviewed more Xilinx Zynq model patches * looking at BE8 support: Paul Brook has posted some patches to support this in user mode * LP:944645: fixed bug where we weren't clearing the IT bits when entering an M profile exception handler * sent out an arm-devs.next pullreq * trying to track down why linux-user is failing brk() and thus causing bash segfaults

13 years, 5 months

1
0
0 0

[RFC] NEON vs. ARM register selection

by Andrew Stubbs

Hi All, As you know, the compiler currently has difficulties choosing between whether to do an operation in NEON or not. As I see it there are three problems: 1. Simply, is it profitable? NEON can do many DImode operations in one or two instructions where 2 to 10 normal ARM/Thumb instructions would be required (not to mention the added register pressure), but there is a cost associated with moving the inputs to NEON, and the results back. If the data can stay in NEON for more than one operation, then that's even better. If the data must be loaded from memory, and the result stored back to memory, then it's only a question of whether the register space is available, or not. Currently these decisions are made in the IRA/reload passes. 2. Values that originate in hard-registers stay there. This applies to function parameters, mostly, but also in general where the result of an operation is allocated first. If there is no instruction that can use the value there then the value is 'reloaded' to a more suitable register. If there is any alternative that avoids the move then the register allocator will use it, regardless of the relatives costs of the other alternatives. This problem is reduced where an operation and move can happen in one instruction, but NEON instructions do not do this much. We can write insns that appear to do it, but these output multiple instructions (see my recent core-SI=>NEON-DI extend patch). 3. It all happens too late. The decision whether to use NEON or not is not made until register allocation time. Naturally this means that most of the optimization passes are already completed. Part of the problem is that the operation almost certainly needs splitting (into whatever form was chosen) and this might not be straight forward, post-reload. (However, the split1 pass is already quite late, so perhaps this isn't such a big deal.) Another part of the problem is that passes such as the two lower-subreg passes make assumptions about the register width which are not accurate if the operation is to end up in NEON. There are other, lesser problems, such as it being hard to adjust the costs for different cores (A8 in particular) and the cost of generating an immediate constant can't be known until it's known what instructions will be used to generate it. These problems are not specific to NEON, of course. I believe IWMMXT suffers from the same issues. Likewise the C6X port, and also the i386 MMX to some degree. Anything that has instructions that only operate on a subset of registers, basically. So, Bernd has suggested an outline of a solution. I've quizzed him on this, added a few of my own ideas, and probably a good selection of misunderstandings, bad assumptions, and general cock ups, and come up with something I can write here for comment. I can post something to upstream later if it doesn't get totally shot down now. The basic idea is that we add a new RTL optimization pass (or two) that assesses the usage of pseudo registers, and makes recommendations about what register class each should end up in, if there's a choice. These recommendations would then be used by later passes to get a better use of NEON. I might call this the "prealloc" pass, or something. Firstly, for each pseudo-register in a function, the pass would look at the insn constraints for each "def" and "use", and see how the registers relate to one another. This might determine things like "if rN is in class A, then rM must be also in class A". E.g. if you have two registers with constraints like this: "r,w" "r,w" .. (and 'r' and 'w' do not overlap) then you know that there is a choice between one mode or another, whereas this: "r,w,r,w" "r,w,w,r" .. would impose no restrictions and we can carry on as normal. Having done that we'd end up with sets of pseudo-registers that must make a decision one way or the other, and we'd know where the operations are that would force a move from one class to the other. There's a fair amount of handwavium in there at present, because I've not worked out what to do with overlapping register classes (think VFP_LO_REGS) and all the other complications. Secondly, the pass would consider the costs of each alternative, and store a recommended register class for each pseudo-register in a table somewhere. It would also create new pseudos and insert extra move instructions at the register file boundaries where an existing register would have had split recommendations (this would solve problem 2 above). Again, there's handwavium in "consider the costs". This isn't too hard for size-optimization (assuming the "length" attributes on the insn is correct), but more difficult for speed optimization. Factors to include would be the move costs (here the A8 issues would be addresses) and the relative speeds of the operations in both alternatives. Also, the various possible transition points between the two modes might need some comparisons. Thirdly, the subsequent passes would need to be modified, as would some of the back-end bits and bobs. 1. Lower-subreg would need to detect 'word_mode' based on the recommended register class, not the global value. 2. The many split patterns in the machine description could be adjusted so that, instead of simply conditionalizing on "reload_completed", they split at split1 if that's the best option. (Maybe it would be profitable to insert a new, earlier split pass specifically for this case to take advantage of the likes of combine? I mean, ideally this decision would have been made at expand time, if it could have been?) It might be useful to *not* split too soon, in some cases, so that the register allocator can still make the final decision based on register pressure, and whatever other factors it uses. Of course, the existing late-split option would need to be retained in case the prealloc pass is disabled, in any case. 3. Various passes would have to be taught not to remove seemingly superfluous register moves where they actually move between register classes. 4. Pretty much nothing would need doing to register allocation! The extra moves should make allocation a register pressure management issue, rather than a question of making it work. DImode operations preallocated to core-registers may already have been lowered, one way or the other (by split1) so there's no decision left there, and if no lowering was necessary then that option ought to be obviously cheaper. If it insists on making contrary decisions then it can be taught to use the recommendation as a hint, perhaps? In specific problem cases it would also be possible to use instruction attributes to disable (or strongly discourage) certain alternatives based on the recommended class. 5. The existing 'onlya8'/'nota8' nonsense can be removed. 6. The register move cost can be set correctly for each core. 7. If a constant is destined for a NEON register, most likely, arm_gen_constant can use the NEON immediate rules to determine the cost. There's clearly a lot of thought that needs to go into the pseudo-register scan and decision making logic, but the whole thing doesn't look like it'll boil down to very much code in the end. There's also the question of where to put the pass? Too early and you'd need to put a second one in to reassess the much changed RTL later, and too late and lower-subreg won't be able to use it. It's possible that it might be better to treat it more like the data-flow analysis where it's not actually a stand-alone pass, but rather a tool other passes can use? That might depend how computationally expensive it is. Any thoughts anyone? Might something like this actually work? Would it be worth spending the time on this? Andrew

13 years, 5 months

2
1
0 0

lp:gcc/4.7

by Andrew Stubbs

Hi Matthias, GCC 4.7.0 has branched upstream. SVN trunk is now 4.8. Could you please create lp:gcc/4.7 from the release branch. Thanks Andrew

13 years, 5 months

1
0
0 0

Please pull upstream rev. 184603 into gcc-linaro 4.7

by Bernhard Rosenkränzer

Hi, 184603 fixes an ICE we're running into with Android test builds. Please pull it in ASAP so I don't have to mess with the CFLAGS as a workaround. ttyl bero

13 years, 5 months

2
2
0 0

Cross toolchain for lucid updated

by Marcin Juszkiewicz

Linaro cross toolchain backport PPA [1] got updated to recent Ubuntu 12.04 versions: - gcc 4.5.3 - gcc 4.6.2 - binutils 2.22 - eglibc 2.15 - linux 3.2 I tested both compiler versions against few packages, U-Boot and Linux 2.6.32. Those packages may also work under Ubuntu 10.10 'maverick' and 11.04 'natty' but it was not tested. Users of Ubuntu 11.10 'oneiric' may install libmpfr1ldbl package [2] and then use 'lucid' repository to install cross compiler while I will prepare packages for their version of Ubuntu. Please report any problems with those backported packages in launchpad [3][4][5]. 1. https://launchpad.net/~linaro-maintainers/+archive/toolchain 2. http://packages.ubuntu.com/lucid/libmpfr1ldbl 3. https://launchpad.net/ubuntu/+source/gcc-4.5-armel-cross/+bugs 4. https://launchpad.net/ubuntu/+source/gcc-4.5-armel-cross/+bugs 5. https://launchpad.net/ubuntu/+source/armel-cross-toolchain-base/+bugs

13 years, 5 months

1
1
0 0

Status call number is unchanged

by Michael Hope

No news yet on the new conference call numbers. Let's see if the old numbers still work. -- Michael

13 years, 5 months

1
0
0 0

Spec 2000 broken?

by Andrew Stubbs

Hi Michael, I ran the benchmark tests on GCC 4.7, but the spec2000 run seems to have failed. This is both on the baseline, and on my patch. See here: http://ex.seabright.co.nz/benchmarks/gcc-linaro-4.7%2bbzr114968~ams-codesou… Andrew

13 years, 6 months

2
2
0 0

Saturated operations

by Michael Hope

Hi Ulrich. The saturated work you're doing at the moment - is it the saturated add/subtract QADD/QSUB/QDADD/QDSUB or the ARMv6 saturate-word-with-shift SSAT/USAT? I've updated the old blueprint and added a new one for SSAT/USAT: https://blueprints.launchpad.net/gcc-linaro/+spec/armv6-saturation-instruct… https://blueprints.launchpad.net/gcc-linaro/+spec/armv5-saturated-ops and was wondering which to put against your name. -- Michael

13 years, 6 months

2
1
0 0

Fleshing out the QEMU blueprints

by Michael Hope

Hi Peter, Rusty. I've fleshed out the KVM blueprints a bit more at: https://wiki.linaro.org/MichaelHope/Sandbox/Q1.12ToDo Notable is adding validation and de-PENDING a few things. I need to know the following things before talking to Validation and Platforms about things that are in their queue: * What's the host kernel? Christoffers, or a linux-linaro-kvm based off his plus work in progress? * What's the guest kernel? * Anything special that needs to be in the host rootfs outside qemu and libvirt? Do you have a host and guest kernel .config that I can pass on to Platforms? -- Michael

13 years, 6 months

1
0
0 0

[ACTIVITY] Feb 20 - Feb 24

by Thiago Jung Bauermann

Hi, GDB on Android: * Set up cross compiling environment using both the Android toolchain and the Linaro toolchain. * Tested Android and Mozilla GDB running as a native binary on the Linaro Android image on qemu. None of them were able to see all threads in a multi-threaded process. The same happens with a native binary on Android talking to a gdbserver on localhost. * I was able to produce an Android GDB 7.1.x binary which when running on an i386 host talking to the Android SDK emulator sees all threads. It is not able to generate a backtrace though. * Carnival holidays. -- []'s Thiago Jung Bauermann Linaro Toolchain Working Group

13 years, 6 months

1
0
0 0

[ACTIVITY] 20th - 24th February

by Andrew Stubbs

Investigated and produced and patch for bug lp936863. Continued work on 64-bit shifts. I updated my neon shifts patch twice: once to take -Os optimization into account, and once because I noticed that the CC clobbers were being retained even after they were known to be not required, and presumably this could be a bad thing. Posted a patch to improve SI->DImode sign- and zero-extends that also move values from core registers to neon registers. Wrote a patch to implement negation in neon register, albeit at the cost of a scratch register to hold the constant zero. This worked, but insisted on loading the zero from memory, despite the fact that NEON has a suitable instruction for loading zero. Attempted to get "vmov dN, #0" to work. The problem appears to be that DImode loads are handled by the VFP load pattern in vfp.md and that that doesn't seem to DTRT for integer constants. I've got as far as trying to figure out what it does do, but there's more to be done to get to the bottom of the problem. Tried to get the benchmarking going for 64-bit shifts. Unfortunately ursa1 has been experiencing outages. Michael has been trying to fix it, but so far I've not been able to run anything there. Instead, I've launched benchmark runs via Michael's normal test infrastructure. Hopefully this can replace the manual ones anyway. That would be more convenient.

13 years, 6 months

1
0
0 0

Running pre-built linaro toolchain failed on Ubuntu 11.04, 64bit server

by Lv Terry-R65388

Hi, I'm trying to use pre-built version of linaro toolchain for cross-compiler in Ubuntu 11.04 on our 64bit server. I got it from http://people.linaro.org/~michaelh/incoming/binaries/. When I run arm-linux-gnueabi-gcc to compile a c source, it says "No such file or directory". The steps are as below: 1. Unpack gcc-linaro-arm-linux-gnueabi-2011.12-20111219+bzr2309~linux.tar.bz2. 2. Rename gcc-linaro-arm-linux-gnueabi-2011.12-20111219+bzr2309~linux to arm-fsl-linux-gnueabi. 3. Copy to ~/toolchain_ltib/gcc-linaro-4.6.3-glibc-2.13-singlelib-2011.12/ 2. Use arm-linux-gnueabi-gcc to compile my code. The output is: r65388@shlinux3:~/MEMCPYBM$ ~/toolchain_ltib/gcc-linaro-4.6.3-glibc-2.13-singlelib-2011.12/arm-fsl-linux-gnueabi/bin/arm-linux-gnueabi-gcc -bash: /home/r65388/toolchain_ltib/gcc-linaro-4.6.3-glibc-2.13-singlelib-2011.12/arm-fsl-linux-gnueabi/bin/arm-linux-gnueabi-gcc: No such file or directory I have installed lsb 4.0 and libc6 2.13. r65388@shlinux3:~/MEMCPYBM$ dpkg -s lsb Package: lsb Status: install ok installed Priority: extra Section: misc Installed-Size: 48 Maintainer: Ubuntu Developers <ubuntu-devel-discuss(a)lists.ubuntu.com> Architecture: all Version: 4.0-0ubuntu16 ... r65388@shlinux3:~/MEMCPYBM$ dpkg -s libc6-dev Package: libc6-dev Status: install ok installed Multi-Arch: same Priority: optional Section: libdevel Installed-Size: 11888 Maintainer: Ubuntu Developers <ubuntu-devel-discuss(a)lists.ubuntu.com> Architecture: amd64 Source: eglibc Version: 2.13-20ubuntu5 Provides: libc-dev Depends: libc6 (= 2.13-20ubuntu5), libc-dev-bin (= 2.13-20ubuntu5), linux-libc-dev Recommends: gcc | c-compiler Suggests: glibc-doc, manpages-dev Breaks: binutils (<< 2.20.1-1), binutils-gold (<< 2.20.1-11), cmake (<< 2.8.4+dfsg.1-5), gcc-4.4 (<< 4.4.6-3ubuntu1), gcc-4.4-base (<< 4.4.6-3ubuntu1), gcc-4.5 (<< 4.5.3-1ubuntu2), gcc-4.5-base (<< 4.5.3-1ubuntu2), gcc-4.6 (<< 4.6.0-12), gcj-4.4-base (<< 4.4.6-2ubuntu2), gcj-4.5-base (<< 4.5.3-1ubuntu2), gnat-4.4-base (<< 4.4.6-1ubuntu3), libhwloc-dev (<< 1.2-3), libjna-java (<< 3.2.7-4), liblouis-dev (<< 2.3.0-2), liblouisxml-dev (<< 2.4.0-2), make (<< 3.81-8.1), pkg-config (<< 0.26-1) ... Do you know why? How can I fix this issue? Thanks~~ Yours Terry

13 years, 6 months

3
25
0 0

[Activity] Week 8

by Zhenqiang Chen

Summary: * Linaro binary toolchain 2012.02 release. * Try the Multilib patch on linaro toolchain. Details: 1. Linaro binary toolchain 2012.02 release. * Create linaro license file based on embedded toolchain, add crosstool-ng and sysroot (eglibc) license. * Tests on win7, oneiric and RHEL5. 2. Try to apply the multilib patch used in embedded toolchain to linaro toolchain. * The multilib solution can not work together with the multiarch patch. (1) With both patches, "gcc -print-multi-lib" can only print one option. (2) With only multilib patch, "gcc -print-multi-lib" can print the multilib list. * Successfully build a multilib toolchain (Based on the prebuilt oneiric-sysroot with directory structure change for include and lib.) without the multiarch patch and c++ support. Plans: 1. Try crosstool-ng 1.14 to build multilib eglibc. 2. Investigate how to make multiarch and multilib work together. Planed leaves: * Mar. 1-2. Best regards! -Zhenqiang

13 years, 6 months

1
0
0 0

KVM blueprints

by Michael Hope

Hi Peter, Rusty. I've written down my notes from Connect and drafted the blueprints at: https://wiki.linaro.org/MichaelHope/Sandbox/Q1.12ToDo I've put my priories on things and sorted them by relative importance. I'll check tomorrow if there's a good match between this, the KVMEpic spec, and the older task spreadsheet. Most of these are too short. I don't know what's involved with the CP15 work. Some still have big PENDINGS. Please edit and correct. There's a lot of work there. Some is on other teams but I'm concerned. -- Michael

13 years, 6 months

1
0
0 0

[ACTIVITY] Feb 20 - Feb 24

by Ulrich Weigand

== GCC == * Merged Ira's remaining vectorizer patches into Linaro GCC 4.7. * Merged Richard's sched-pressure patch (enabled by default on ARM) into Linaro GCC 4.7. Asked IBM colleagues for benchmark results on s390 and rs6000 as well. * Committed patch to enable vect_condition tests in FSF mainline, FSF 4.6, and Linaro GCC 4.7. * Verified that Linaro GCC 4.7 now contains every feature that was present in Linaro GCC 4.6. * Rebased and re-benchmarked fwprop-subreg patch. * Ongoing work on a patch to generate usat/ssat instructions where appropriate. * Worked on improving end-of-loop value computation. * Fixed mainline .init_array/.fini_array regression on certain newlib targets. * Reviewed several merge requests. Mit freundlichen Gruessen / Best Regards Ulrich Weigand -- Dr. Ulrich Weigand | Phone: +49-7031/16-3727 STSM, GNU compiler and toolchain for Linux on System z and Cell/B.E. IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter | Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen | Registergericht: Amtsgericht Stuttgart, HRB 243294

13 years, 6 months

1
0
0 0

[Activity] Week 8

by Ramana Radhakrishnan

On holiday last week and at connect week before that. ==Progress=== * Recovered from jet-lag and started to catch up on email. * Patch review week. * Cleared out a bit of my patch backlog. * Helped Asa with some bug triaging. * Read up a bit about ssat and usat and what we should be doing with other tuning for them. * Looked up some performance profiles with v8 benchmarks . Reviewed the Android benchmark numbers. * Booked tickets for connect 2012.q2 * Wrote up connect week discussions on libav regressions . === Plans === * Clear out more pending patches ( PGO next on the hit list and the vmla investigation) * Investigate partial-partial PRE results. * Finish off the ABI tests. Absences. * 1 week holiday sometime before that - to be booked. * Linaro Connect Q2.12 - May 28 - June 1

13 years, 6 months

1
0
0 0

[ACTIVITY] weekly status

by Ken Werner

Hi, OpenEmbedded: * added initial support for the Linaro GCC 4.6 based toolchain to the meta-linaro layer. * allows to build the Linaro toolchain the OE way * successfully built the core-image-sato but when running inside QEMU the GUI isn't usable (needs investigation) glibc: * built glibc+glibc-ports natively on ARM * the upstream glibc testsuite isn't aware of the multiarch paths and therefore it won't find some libs (libgcc_s, libstdc++) * workaround using symlinks; wiki updated: https://wiki.linaro.org/KenWerner/Sandbox/HowToBuildToolchainComponents#gli… Regards, Ken

13 years, 6 months

1
0
0 0

Re: gcc: Thumb interworking and weakly linked functions

by V, Aneesh

+ linaro-toolchain Hello Ulrich, I want to revisit this old thread. Sorry for the sloppy follow-up. But, this time around I have more data. On Tue, Mar 15, 2011 at 9:00 PM, Ulrich Weigand <Ulrich.Weigand(a)de.ibm.com> wrote: > Aneesh V <aneesh(a)ti.com> wrote: > >> I was trying to build u-boot in Thumb2 for OMAP4. Everything was fine >> until I added some patches recently. One of these patches introduced an >> API (let's say foo()) that has a weakly linked alias(let's say >> __foo()) and a strongly linked implementation(the real foo()) in an >> assembly file. >> >> Although I give -mthumb and -mthumb-interwork for all the files, >> apparently GCC generates ARM code for assembly files. In the final >> image foobar() calls foo() using a BL. Since foobar() is in Thumb and >> foo() in ARM, it ends up crashing. Looks like foobar() assumed foo() >> to be Thumb because __foo() is Thumb. > > I'm unable to reproduce this. Do you have a complete test case? > > I've tried with the following small example: > > foo1.c: > > extern void foo (void) __attribute__ ((weak, alias ("__foo"))); > > void __foo (void) > { > } > > int main (void) > { > foo (); > } > > foo2.S: > .text > .align 2 > .global foo > .type foo, %function > foo: > push {r7} > add r7, sp, #0 > mov sp, r7 > pop {r7} > bx lr > .size foo, .-foo > > When building just "gcc foo1.c", I get: > > 0000835c <__foo>: > 835c: b480 push {r7} > 835e: af00 add r7, sp, #0 > 8360: 46bd mov sp, r7 > 8362: bc80 pop {r7} > 8364: 4770 bx lr > 8366: bf00 nop > > 00008368 <main>: > 8368: b580 push {r7, lr} > 836a: af00 add r7, sp, #0 > 836c: f7ff fff6 bl 835c <__foo> > 8370: 4618 mov r0, r3 > 8372: bd80 pop {r7, pc} > > When building both files "gcc foo1.c foo2.S", I get instead: > > 00008368 <main>: > 8368: b580 push {r7, lr} > 836a: af00 add r7, sp, #0 > 836c: f000 e802 blx 8374 <foo> > 8370: 4618 mov r0, r3 > 8372: bd80 pop {r7, pc} > > 00008374 <foo>: > 8374: e92d0080 push {r7} > 8378: e28d7000 add r7, sp, #0 > 837c: e1a0d007 mov sp, r7 > 8380: e8bd0080 pop {r7} > 8384: e12fff1e bx lr > > > So it seems to me the linker is handling this correctly ... > > (This is on Ubuntu Natty using system gcc and binutils.) I could reproduce the problem on older tool-chain(Sourcery G++ Lite 2010q1-202) [1] with a modified version of your sample code: a.c: ==== extern void foo (void) __attribute__ ((weak, alias ("__foo"))); void __foo (void) { } extern void call_foo(void); int main (void) { call_foo (); } b.S: ==== .text .align 2 .global foo foo: push {r7} add r7, sp, #0 mov sp, r7 pop {r7} bx lr .size foo, .-foo c.S .text .align 2 .global call_foo call_foo: bl foo bx lr .global __aeabi_unwind_cpp_pr0 __aeabi_unwind_cpp_pr0: bx lr Now, I build them using the following commands, which is similar to what U-Boot does: arm-none-linux-gnueabi-gcc -mthumb -mthumb-interwork -c a.c arm-none-linux-gnueabi-gcc -mthumb -mthumb-interwork -c b.S arm-none-linux-gnueabi-gcc -mthumb -mthumb-interwork -c c.S arm-none-linux-gnueabi-ld -r a.o -o alib.o arm-none-linux-gnueabi-ld -r b.o -o blib.o arm-none-linux-gnueabi-ld -r c.o -o clib.o arm-none-linux-gnueabi-ld --start-group clib.o alib.o blib.o --end-group -o a.out armobjdump -S --reloc a.out You will get something like: 00008094 <call_foo>: 8094: fa000006 blx 80b4 <foo> 8098: e12fff1e bx lr Please note that that the 'blx' is not correct. Now, do the following change: diff --git a/b.S b/b.S index e0f2de9..96dba1f 100644 --- a/b.S +++ b/b.S @@ -1,5 +1,6 @@ .text .align 2 +.type foo, %function .global foo foo: push {r7} And build it again the same way and you will see: 00008094 <call_foo>: 8094: eb000006 bl 80b4 <foo> 8098: e12fff1e bx lr I can't reproduce this on Linaro GCC 2012.01, so looks like the problem is solved in recent tool-chains. However, sadly I could reproduce a different but similar problem with Linaro GCC 2012.01. This time the call is from C(Thumb) to assembly(ARM) and no weakly linked symbols are involved. a.c: ==== int main (void) { foo (); } b.S: ==== .text .align 2 .global foo foo: push {r7} add r7, sp, #0 mov sp, r7 pop {r7} bx lr .size foo, .-foo .global __aeabi_unwind_cpp_pr0 __aeabi_unwind_cpp_pr0: bx lr arm-linux-gnueabi-gcc -mthumb -mthumb-interwork -c a.c arm-linux-gnueabi-gcc -mthumb -mthumb-interwork -c b.S arm-linux-gnueabi-ld -r a.o -o alib.o arm-linux-gnueabi-ld -r b.o -o blib.o arm-linux-gnueabi-ld --start-group alib.o blib.o --end-group -o a.out arm-linux-gnueabi-objdump -S --reloc a.out gives: 8076: af00 add r7, sp, #0 8078: f000 f802 bl 8080 <foo> 807c: 4618 mov r0, r3 It should have been "blx 8080 <foo>", isn't it? Again, %function solves it. I agree that not marking the assembly functions ' %function' is a problem in the code, so it's not a critical bug. But I would've been happier if the linker refused to link it rather than branching with the wrong instruction. Isn't that a problem? Problem No:2 ************* Linaro GCC 2012.01 is giving a new problem w.r.to Thumb build that is not existing in Sourcery G++ Lite 2010q1-202. However, I couldn't reproduce this problem with a small program like above. So, let me give you reference to the original u-boot code that shows the problem and steps to reproduce it. tree: git://github.com/aneeshv/u-boot.git branch: thumb The above branch has mainline u-boot with 4 additional patches from me for enabling Thumb build. You can build it like this: make CROSS_COMPILE=arm-linux-gnueabi- ARCH=arm distclean make CROSS_COMPILE=arm-linux-gnueabi- ARCH=arm omap4_sdp4430 This builds two images u-boot and u-boot-spl. SPL is a tiny u-boot that runs from internal RAM and loads the u-boot to SDRAM. Now, please have a look at the map file of u-boot-spl which is at: spl/u-boot-spl.map I see the following in my map file: /spl/u-boot-spl.map: ================= .rodata.wkup_padconf_array_essential_4460 0x40309583 0x4 board/ti/sdp4430/libsdp4430.o 0x40309583 wkup_padconf_array_essential_4460 .rodata.wkup_padconf_array_essential 0x40309587 0xc board/ti/sdp4430/libsdp4430.o 0x40309587 wkup_padconf_array_essential .rodata.core_padconf_array_essential 0x40309593 0x60 board/ti/sdp4430/libsdp4430.o 0x40309593 core_padconf_array_essential Please note that the .rodata symbols have odd addresses. These arrays actually need to be aligned at least to half-word boundary. In fact, in the image I verified that they are put at even addresses. So, the symbols have been kept as real address +1. I understand that this is the convention for Thumb functions. I guess the tool-chain did it for data too?? And I am getting unaligned access aborts on accessing them. I notice that this doesn't happen with all .rodata. symbols in the image. I couldn't see any difference between working and non-working files nor any difference in the command used to build them! Well, this doesn't happen if I don't use "-fdata-sections" in gcc options. So, apply the following patch and you will see that those symbols have even addresses now. diff --git a/config.mk b/config.mk index ddaa477..723286a 100644 --- a/config.mk +++ b/config.mk @@ -190,7 +190,7 @@ CPPFLAGS := $(DBGFLAGS) $(OPTFLAGS) $(RELFLAGS) \ # Enable garbage collection of un-used sections for SPL ifeq ($(CONFIG_SPL_BUILD),y) -CPPFLAGS += -ffunction-sections -fdata-sections +CPPFLAGS += -ffunction-sections LDFLAGS_FINAL += --gc-sections endif /spl/u-boot-spl.map: ================= .rodata 0x40309204 0x38c board/ti/sdp4430/libsdp4430.o 0x40309204 core_padconf_array_essential 0x40309264 wkup_padconf_array_essential 0x40309270 wkup_padconf_array_essential_4460 0x40309274 core_padconf_array_non_essential 0x40309540 wkup_padconf_array_non_essential 0x40309588 wkup_padconf_array_non_essential_4430 Will you be able to look into these? Thanks, Aneesh [1] Sourcery G++ Lite 2010q1-202 arm-none-linux-gnueabi-gcc (Sourcery G++ Lite 2010q1-202) 4.4.1 GNU ld (Sourcery G++ Lite 2010q1-202) - binutils 2.19.51.20090709 [2] Linaro 4.6-2012.01 arm-linux-gnueabi-gcc (crosstool-NG linaro-1.13.1-2012.01-20120125 - Linaro GCC 2012.01) 4.6.3 20120105 (prerelease) GNU ld (crosstool-NG linaro-1.13.1-2012.01-20120125 - Linaro GCC 2012.01) 2.22

13 years, 6 months

6
11
0 0

[ACTIVITY] w08

by Asa Sandahl

Hi! * Development benchmarks: Focused on the implementation of development benchmarks in the cbuild system. Discussed with Michael about how the comments should look. I am now testing my solution in the launchpad staging area. You can see a link to an example of a benchmark result comment below. (Should not be a double comment of course, and the actual result files are only stored on my machine for now.) https://code.staging.launchpad.net/~ramana/gcc-linaro/partial_partial_pre_t… The real benchmark names are not mentioned, since that might be sensitive. * Bug reports I got an action on Monday's meeting to investigate if https://bugs.launchpad.net/ubuntu/+source/gcc-4.6/+bug/926855 is the same as PR52294. Investigated with help from Ramana and updated the ticket. Regards Åsa

13 years, 6 months

1
0
0 0

[ACTIVITY] report week 08

by Peter Maydell

Current Milestones: || || Planned || Estimate || Actual || ||cp15-rework || 2012-01-06 || 2012-??-?? || || (new blueprints & reestimate for this one pending Michael H's writeup/etc) Historical Milestones: ||a15-usermode-support || 2011-11-10 || 2011-11-10 || 2011-10-27 || ||upstream-omap3-cleanup || 2011-11-10 || 2011-12-15 || 2011-12-12 || ||initial-a15-system-model || 2012-01-27 || 2012-01-27 || 2012-01-17 || ||qemu-kvm-getting-started || 2012-03-04?|| 2012-03-04?|| 2012-02-01 || == cp15-rework == * figuring out how various suggestions/conversations on this from Connect ought to fit together == other == * LP:928580 recent u-boot not booting on QEMU beagle: tracked this down to a u-boot bug and passed to John Rigby * investigated a linux-user segfault deadlock which turns out to be because the guest program is using setrlimit to restrict memory usage and one of qemu's internal memory allocations then fails. Not clear how best to fix but was able to nack the suggested patch... * most of the patches for passing a device tree to the kernel from qemu (via -dtb option) are now upstream; rebased the final patch which should now go in soon I think * LP885239 qemu-linaro broke booting of zaurus models: fixed a bug where a trustzone support patch had broken some cp15 registers for xscale cpus * minor patch: fixed formatting of list of supported machines help text * fix an embarrassing segfault bug in a patch of mine committed last week * it turns out that we can detect at compile time whether glibc's makecontext() is an always-fails stub : sent configure patch to do this * travel mostly set up for connect Q2.12 * lots of meetings/training: toolchain call; standup; 1-2-1; project management training; comms meeting

13 years, 6 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

linaro-toolchain