[gem5-dev] Re: Adding an extension object to Packet-s

This is to mimic systemc's protocol extension mechanism, where you can add side channel info to a transaction. Sender state is more for keeping track of information the sender needs related to a particular packet, rather than to send additional state down the line to other consumers. For instance,

[gem5-dev] Adding an extension object to Packet-s

Hi folks. Yan Lee, a Google colleague of mine, is interested in adding one or more extension objects to gem5 Packet. Systemc has a mechanism like this already, and it's come up a few times where we've wanted to have something like this on the gem5 side as well. Yan, can you please talk about the

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #189

https://gem5-review.googlesource.com/c/public/gem5/+/58633 On Tue, Apr 5, 2022 at 7:52 PM Jason Lowe-Power wrote: > That sounds like a great middle ground. > > We could *probably* drop support for 18.04 in v22.0, but we should have a > broader discussion and make sure that's our intention. It's

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #189

Well, in that case, we could detect version 3.0.1 (there is a function to get the version I believe) and do the monkey patch they suggest in the email. The fix was just adding that little function they show, which did not exist at all in version 3.0.1. I would have to look up exactly how you

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #189

The bug in SCons was fixed in version 3.0.2. I have a CL which updates the required version here: https://gem5-review.googlesource.com/c/public/gem5/+/58629 I'm assuming some docker configuration something will need to be updated to use SCons version 3.0.2 instead of the 3.0.1 it seems to be

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #189

I think this is actually a bug in SCons. This docker image seems to be using the minimum version of it we support, which I think is where the problem is coming from and not the compiler versions. There is a "fix"/workaround in this email thread which I haven't tried yet, and which is a bit of a

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #189

Yes, I'll take a look. Gabe On Mon, Apr 4, 2022 at 3:02 PM Bobby Bruce wrote: > I don't fully understand the issue, but after doing a bisect, the commit > causing these issues is > https://gem5-review.googlesource.com/c/public/gem5/+/58356. > > To reproduce locally: > > ``` > docker run -u

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #187

It was a little different, but still pretty simple. https://gem5-review.googlesource.com/c/public/gem5/+/58349 Gabe On Tue, Mar 29, 2022 at 6:00 PM Gabe Black wrote: > My guess is that I missed add a ['CONF'] in front of one of one of the > uses of a configuration variable. No requirements

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #187

My guess is that I missed add a ['CONF'] in front of one of one of the uses of a configuration variable. No requirements should have changed, so this is likely a simple bug where some bookkeeping is getting dropped somewhere. I'll take a look. Gabe On Tue, Mar 29, 2022 at 8:44 PM Bobby Bruce

[gem5-dev] Re: Build failed in Jenkins: nightly #161

https://gem5-review.googlesource.com/c/public/gem5/+/57749 On Tue, Mar 15, 2022 at 12:59 PM Gabe Black wrote: > It looks like the problem is that the aapcs test calls ThreadContext which > returns a TheISA::VecRegContainer, where since that test is for ARM it > *should* really be working with

[gem5-dev] Re: Build failed in Jenkins: nightly #161

It looks like the problem is that the aapcs test calls ThreadContext which returns a TheISA::VecRegContainer, where since that test is for ARM it *should* really be working with an ARM VecRegContainer. Because this is being built for NULL, that uses the dummy versions which are not complete enough

[gem5-dev] resource downloader failure

Hi folks. I've been trying to run tests locally, and I've been running into occasional flakiness due to a problem with the resource downloader. This reminds me somewhat of flakiness I was seeing in kokoro as well, so they are probably related. The error output is this: $ cat

[gem5-dev] microcode test system

Hey testing folks (primarily Bobby probably?). I'm revamping the x86 microcode to extract it from SCons itself, and also just to make it less painful and annoying to work with. Importantly, the microcode basically lists a sequence of microop mnemonics and a set of arguments. The mnemonic simply

[gem5-dev] Re: build configuration overhaul

variables. Similarly, we could > remove the common build directory. These things would solve some of the > problems you enumerate without breaking compatibility. It's definitely > low-hanging fruit. > Yes, exactly, that's my plan. > > Cheers, > Jason > > > &

[gem5-dev] build configuration overhaul

Hi folks. I have a design document here: https://docs.google.com/document/d/11KspjOClPzGRqLUgNTfKQPrrkmsqx-2ulC71nZUJ2qA/edit?usp=sharing which describes my proposal to overhaul how gem5 builds are configured. Please take a look and provide feedback. Important points to note are (optionally but

[gem5-dev] Re: For fun x86 bare metal CLs

BTW, I've started trying to get this test program to run to flush out bugs in our pre-64bit x86 support. It's been very helpful and usable, and I'm hoping once I get it working (or at least mostly working :-) we can add it as a regression. https://github.innominds.com/barotto/test386.asm Gabe

[gem5-dev] Re: For fun x86 bare metal CLs

It seems slow, and I had to fix a bug in SeaBIOS, but it seems to run :-) No video currently, since I haven't written a VGA device yet. Gabe On Tue, Jan 18, 2022 at 9:03 PM Gabe Black wrote: > HimemX 3.34 [Sep 05 2015] (c) 1995, Till Gerken 2001-2006 Tom Ehlert > Always on A20 method used >

[gem5-dev] Re: For fun x86 bare metal CLs

HimemX 3.34 [Sep 05 2015] (c) 1995, Till Gerken 2001-2006 Tom Ehlert Always on A20 method used Kernel: allocated 45 Diskbuffers = 23940 Bytes in HMA FreeCom version 0.84-pre2 XMS_Swap [Aug 28 2006 00:29:00] Done processing startup files C:\FDCONFIG.SYS and C:\AUTOEXEC.BAT Type HELP to get

[gem5-dev] Incorrect handling of PCI interrupt configuration

Hi folks. I was looking at how our PCI system is set up in gem5, and I've found that the way interrupts is configured is not correct. You can find a fairly complete description of the problem over here: https://gem5.atlassian.net/browse/GEM5-1165 My plan to fix this is to move the description of

[gem5-dev] For fun x86 bare metal CLs

Hey folks, I've wanted to try to get a bare metal x86 config going for a long time, since I thought it would be fun to play some old DOS games from my childhood on gem5. To that end, I decided to try to get SeaBIOS, a free implementation of a PC BIOS, to run there. While I haven't gotten that to

[gem5-dev] Re: Broken SST due to python changes

I just realized I turned off my computer since I'll be traveling, and my test run of sst can't finish if the computer is off :-P. I have not completed a run of your example command, but it starts and I have no reason to believe it won't finish. I would still suggest running it yourself (with my

[gem5-dev] Re: Broken SST due to python changes

https://gem5-review.googlesource.com/c/public/gem5/+/54325/1 On Wed, Dec 15, 2021 at 11:12 PM Gabe Black wrote: > There will be a couple more patches coming, since the systemc subsystem > has problems when python is already running when static initializers run. I > had been building without

[gem5-dev] Re: Broken SST due to python changes

There will be a couple more patches coming, since the systemc subsystem has problems when python is already running when static initializers run. I had been building without systemc enabled to simplify things earlier, but I need to get it running too. That should be a fairly quick fix. Gabe On

[gem5-dev] Re: Broken SST due to python changes

I have it running, although it's taking a while. How long should it take to finish? Should it finish immediately, 15 minutes, a few hours, tomorrow...? I'll upload my changes in the mean time. Gabe On Wed, Dec 15, 2021 at 9:13 PM Gabe Black wrote: > Ok, I did/am doing a little more looking,

[gem5-dev] Re: Broken SST due to python changes

Ok, I did/am doing a little more looking, and part of the problem seems to be that some of the python blobs executed with executePython assume prior blobs have done imports for them already. That was where the modules or their contents not being found was coming from, and then after that something

[gem5-dev] Re: Broken SST due to python changes

Also, I have to say, while I'm grateful that there was a push to get the SCons changes reviewed for the release, we wouldn't be in this crunch if they had been reviewed a month ago. It would be best not to let things bunch up and then make a herculean push right at the deadline, where problems

[gem5-dev] Re: Broken SST due to python changes

That's actually very helpful, since I was pretty sure there was a memory problem but couldn't figure out where. I think there's a good chance this will be an easy fix. Gabe On Wed, Dec 15, 2021, 8:59 AM Jason Lowe-Power wrote: > Ok, I was able to get a debugger to work and I dug in a bit

[gem5-dev] Re: Broken SST due to python changes

https://gem5-review.googlesource.com/c/public/gem5/+/54006 At least one problem I ran into was that the order of static initializers was not determinstic, so the structures which said what embedded python modules existed might be statically constructed after the thing that consumes and sets them

[gem5-dev] Re: Broken SST due to python changes

Thanks, that was really helpful. I've fixed the compilation problems, and then a few more that were hiding behind that, including a few ones that were intrinsic to the way sst was initializing the gem5 library (with dlopen). Was this working to begin with and all this breakage is new, or was it

[gem5-dev] Re: Broken SST due to python changes

Dumb question: I'm trying to run gdb inside this container on the sst thing. How do I do that? It's not installed in the container now, and I can't (easily) figure out how to get it installed. I can tell docker to install it, but then it seems to throw that away as soon as the command ends. Gabe

[gem5-dev] Re: Broken SST due to python changes

Hi Bobby, not yet, I meant to look into this for the last couple days but kept running out of time. I'm sitting down to work on it right now. Gabe On Fri, Dec 10, 2021 at 1:21 PM Bobby Bruce wrote: > Hey Gabe, > > Is there any update on this? > > Kind regards, > Bobby > -- > Dr. Bobby R. Bruce

[gem5-dev] Re: redundant builds in the long regressions

Ok, great, thanks! On Mon, Dec 6, 2021 at 11:20 AM Bobby Bruce wrote: > Hey Gabe, > > That change is still in review: > https://gem5-review.googlesource.com/c/public/gem5/+/53503 > > I just wanted to get a few more eyes on it before submitting. You are > correct, after this is merged we'll no

[gem5-dev] Re: redundant builds in the long regressions

And also GCN3_X86, which hadn't started yet. On Mon, Dec 6, 2021 at 4:27 AM Gabe Black wrote: > Hi Bobby and other folks. While I'm waiting for reviews on my register > related multiISA patch series, I'm running long regressions on it to try to > preemptively track down any bugs that might be

[gem5-dev] redundant builds in the long regressions

Hi Bobby and other folks. While I'm waiting for reviews on my register related multiISA patch series, I'm running long regressions on it to try to preemptively track down any bugs that might be in it. I notice that the test framework is build both build/X86_MESI_Two_Level/gem5.opt and

[gem5-dev] Re: Build failed in Jenkins: nightly #48

I think this is the problem: scons: `build/NULL/unittests.perf' is up to date. scons: `build/NULL/unittests.prof' is up to date. The prof and perf suffixes are no more. Those probably shouldn't have been using prof or perf anyway, at least not exclusively, because we definitely want asserts,

[gem5-dev] Re: Gem5 on MacOS/clang --as-needed flag detection

Hi Arun, thanks for letting us know about this. Would you mind creating a bug on Jira? https://gem5.atlassian.net/jira/software/c/projects/GEM5/issues I'll have to look at this more closely, but I think this is a bug in the CheckLinkFlag check, where it should not set SHLINKFLAGS even if

[gem5-dev] Re: failing kvm tests

gt;> patch: https://gem5-review.googlesource.com/c/public/gem5/+/52384, and >>> I'm currently looking into the bugs in the very-long tests. >>> >>> Kind regards, >>> Bobby >>> -- >>> Dr. Bobby R. Bruce >>> Room 3050, >>> Kemper Hall,

[gem5-dev] Re: multi-ISA gem5 proof of concept

On Mon, Nov 8, 2021 at 2:13 PM Gabe Black wrote: > > > On Mon, Nov 8, 2021 at 8:40 AM Jason Lowe-Power > wrote: > >> This looks quite promising! Uploading a branch to gerrit is a good idea. >> > > Ok, I'll do that. All but 4 of the changes in this composite branch are > uploaded for review, and

[gem5-dev] Re: multi-ISA gem5 proof of concept

3" It looks like >> current develop also has this problem, but it's a bit worrying. I can't >> figure out why the hello application wouldn't exit with code 0... >> >> Cheers, >> Jason >> >> On Sun, Nov 7, 2021 at 5:38 PM Gabe Black via gem5-dev >&g

[gem5-dev] Re: multi-ISA gem5 proof of concept

because we don't explicitly return anything from main, and so the return value from printf is still in the return value register, which is the number of characters printed. > Cheers, > Jason > > On Sun, Nov 7, 2021 at 5:38 PM Gabe Black via gem5-dev > wrote: > >> Build

[gem5-dev] Re: failing kvm tests

gt; -- > Dr. Bobby R. Bruce > Room 3050, > Kemper Hall, UC Davis > Davis, > CA, 95616 > > web: https://www.bobbybruce.net > > > On Wed, Nov 3, 2021 at 4:22 AM Gabe Black via gem5-dev > wrote: > >> Hey folks, I recently discovered KVM wasn't set up on my desktop, whi

[gem5-dev] Re: multi-ISA gem5 proof of concept

Build performance improvements: Building all 6 ISAs separately. $ time scons build/ARM/gem5.opt build/MIPS/gem5.opt build/POWER/gem5.opt build/RISCV/gem5.opt build/SPARC/gem5.opt build/X86/gem5.opt real37m0.210s user764m20.963s sys 46m18.113s $ du -sh build 16G build

[gem5-dev] multi-ISA gem5 proof of concept

As mentioned in the other thread, I can upload this hash as a branch on gerrit if people want to try it out. Gabe $ git log --oneline origin/develop..multiarch | wc -l 203 $ build/ALL/gem5.opt configs/example/se.py --cpu-type=ArmAtomicSimpleCPU -c tests/test-progs/hello/bin/arm/linux/hello

[gem5-dev] Re: Build failed in Jenkins: nightly #27

>>>>>>> On Wed, Nov 3, 2021 at 12:45 PM Jason Lowe-Power < >>>>>>>>>>>>>> ja...@lowepower.com> wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> Here's my d

[gem5-dev] Re: Build failed in Jenkins: nightly #27

g >>>>>>>>>>>>> on other >>>>>>>>>>>>> things while it runs. >>>>>>>>>>>>> >>>>>>>>>>>>> Gabe >>>>>>>>>>>>> >>>>>>>&

[gem5-dev] Re: Build failed in Jenkins: nightly #27

>>>>>>>>>>>> if/when it >>>>>>>>>>>>> finishes. >>>>>>>>>>>>> >>>>>>>>>>>>> Gabe >>>>>>>>>>>&g

[gem5-dev] Re: Build failed in Jenkins: nightly #27

to do >>>>>>>>>>>>> with >>>>>>>>>>>>> things getting cleaned up and/or destructed at the end of the >>>>>>>>>>>>> simulation, >>>>>>>>&g

[gem5-dev] Re: Build failed in Jenkins: nightly #27

e >>>>>>>>>>>> >>>>>>>>>>>> On Tue, Nov 2, 2021 at 3:46 AM Gabe Black >>>>>>>>>>>> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> A clean build seems t

[gem5-dev] Re: Build failed in Jenkins: nightly #27

;>>>>>>>>>> about not being able to find IdeDisk which is very strange... I >>>>>>>>>>>> don't think >>>>>>>>>>>> I have an account on the machine where this ran, but ideally I'll >>>>

[gem5-dev] default m5ops_base to off on x86

Hey folks. One of the last things to sort out before I'd consider the gem5 mutli-ISA support at least usable is that the m5ops_base config parameter defaults to zero, except if you're on x86 where it defaults to 0x. I think it would make sense to instead default it to 0 all the time,

[gem5-dev] Re: Build failed in Jenkins: nightly #27

will be easier to work with. >>>>>>>>>>> >>>>>>>>>>> Gabe >>>>>>>>>>> >>>>>>>>>>> On Mon, Nov 1, 2021 at 3:03 PM Jason Lowe-Power < >>>>>>>>&

[gem5-dev] Re: preview branch for mutli-ISA gem5?

anch we can test for a >> while, it will be less likely to break our users use cases. >> >> Cheers, >> Jason >> >> On Wed, Nov 3, 2021 at 11:51 PM Gabe Black via gem5-dev < >> gem5-dev@gem5.org> wrote: >> >>> Hey folks, I'm not quite ready for it

[gem5-dev] Re: Build failed in Jenkins: nightly #27

gt;>>>>> and no stats are generated: >>>>>>>>>> ../gem5/> build/ARM/gem5.opt tests/gem5/fs/linux/arm/run.py >>>>>>>>>> tests/gem5/configs/realview-o3.py tests/gem5/resources/arm . >>>>>>>>>

[gem5-dev] preview branch for mutli-ISA gem5?

Hey folks, I'm not quite ready for it yet, but I'm getting close to having everything at least written for a multi-ISA version of gem5. What do you think about having a preview branch on gerrit? I can create one pretty easily, I just want to make sure that's appropriate and something other people

[gem5-dev] failing kvm tests

Hey folks, I recently discovered KVM wasn't set up on my desktop, which I just corrected. Now that it's enabled, the KVM tests have started running, and they are also failing. This could be sort of weird issue on my machine, and I'm currently looking into the failure itself. The thing I wanted to

[gem5-dev] Re: Build failed in Jenkins: nightly #27

;>>>>> I don't think so. The binaries haven't been updated since April >>>>>> ('aarch-system-20210904.tar.bz2'). Well, the blame says April even if the >>>>>> filename is confusing (DDMM?). >>>>>> >>>>

[gem5-dev] Re: Build failed in Jenkins: nightly #27

ch-system-20210904.tar.bz2'). Well, the blame says April even if the >>>>> filename is confusing (DDMM?). >>>>> >>>>> Here's the failure: >>>>> https://jenkins.gem5.org/job/nightly/ws/tests/testing-results/SuiteUID%3Arealview-o3

[gem5-dev] Re: Build failed in Jenkins: nightly #27

.tar.bz2'). Well, the blame says April even if the >>>> filename is confusing (DDMM?). >>>> >>>> Here's the failure: >>>> https://jenkins.gem5.org/job/nightly/ws/tests/testing-results/SuiteUID%3Arealview-o3-ARM-x86_64-opt/TestUID%3Arealview-o3-ARM-x86_64-o

[gem5-dev] Re: Build failed in Jenkins: nightly #27

Arealview-o3-ARM-x86_64-opt/TestUID%3Arealview-o3-ARM-x86_64-opt/simerr >>> >>> build/ARM/arch/arm/insts/pseudo.cc:173: warn: instruction 'csdb' >>> unimplemented >>> build/ARM/arch/arm/generated/decoder-ns.cc.inc:100643: warn: instruction >>> 'mcr bpiall' unimp

[gem5-dev] Re: Build failed in Jenkins: nightly #27

esting-results/SuiteUID%3Arealview-o3-ARM-x86_64-opt/TestUID%3Arealview-o3-ARM-x86_64-opt/simerr >> >> build/ARM/arch/arm/insts/pseudo.cc:173: warn: instruction 'csdb' >> unimplemented >> build/ARM/arch/arm/generated/decoder-ns.cc.inc:100643: warn: instruction >> 'mcr b

[gem5-dev] Re: Build failed in Jenkins: nightly #27

Maybe you need to re-download the resources? Gabe On Sat, Oct 30, 2021 at 3:50 AM jenkins-no-reply--- via gem5-dev < gem5-dev@gem5.org> wrote: > See > > Changes: > > > -- > [...truncated 809.58

[gem5-dev] Re: Build failed in Jenkins: nightly #25

I just ran the long regressions on ToT, and everything passed for me: $ docker run --volume ${PWD}:${PWD} -w ${PWD}/tests --rm gcr.io/gem5-test/ubuntu-20.04_all-dependencies ./main.py run --length long -j24 -t12 [...] Test: realview-o3-ARM-x86_64-opt Passed [...] $ git log -n 1 --oneline

[gem5-dev] Re: Build failed in Jenkins: nightly #25

Also, how do I run these (ideally just the failing test) locally? Gabe On Fri, Oct 29, 2021 at 5:17 PM Gabe Black wrote: > Oh, I didn't see this until now. I'll take a look. > > Gabe > > On Fri, Oct 29, 2021 at 7:45 AM Jason Lowe-Power > wrote: > >> Have you had a chance to look into this,

[gem5-dev] Re: Build failed in Jenkins: nightly #25

Oh, I didn't see this until now. I'll take a look. Gabe On Fri, Oct 29, 2021 at 7:45 AM Jason Lowe-Power wrote: > Have you had a chance to look into this, Gabe? > > On Thu, Oct 28, 2021 at 7:35 AM Jason Lowe-Power > wrote: > >> The realview ARM O3 test is failing. I would guess that it's not

[gem5-dev] Re: CPU complex wrapper

he component library and require > the minimum changes to the (hopefully soon to be deprecated) se/fs.py > config scripts. > > Cheers, > Jason > > On Wed, Oct 27, 2021 at 9:40 PM Gabe Black via gem5-dev > wrote: > >> Hi folks. There are some helper functions in BaseCP

[gem5-dev] CPU complex wrapper

Hi folks. There are some helper functions in BaseCPU which help set up ancillary structures like caches, and tries to keep track of the frontier of ports that need to be connected so that a CPU + caches can be generically hooked into a system. This code is a bit clunky and complex, and makes it

[gem5-dev] Presentation about multi ISA gem5

Hi folks. I wrote a presentation about my work to make gem5 able to support more than one ISA at a time. It's *slightly* out of date now, since more work has been finished since it was last updated. The structure of it seemed like a good idea initially, but I think it may have made the

[gem5-dev] Re: Build failed in Jenkins: nightly #17

I think this is from my change which switched from "if env['TARGET_ISA'] == 'foo'" to using tags. I saw this a while ago, but I've been up literally all night trying to fix problem after problem rebasing my changes on top of ToT, and I haven't had a chance to look at it yet. When I finally get

[gem5-dev] Re: gem5 gui tool?

> CA, 95616 > > web: https://www.bobbybruce.net > > > On Mon, Oct 18, 2021 at 10:47 PM Gabe Black via gem5-dev < > gem5-dev@gem5.org> wrote: > >> If you want to try my script, or even improve it (probably not hard!), >> it's attached. >> >> You

[gem5-dev] Re: gem5 gui tool?

If you want to try my script, or even improve it (probably not hard!), it's attached. You can run it by first collecting a log of timestamps, which I did with a clean build: scons --debug=action-timestamps -j48 build/ARM/gem5.opt | tee log.txt and then run the script: ./scons_timeline.py

[gem5-dev] Re: gem5 gui tool?

Wow, now I remember why I don't like GUI programming :-/ Well, I sort of got some results, and the upshot is that linking gem5 takes a long time :-P. And I am able to at least superficially have 48 things going during the build, although it is not necessarily 48 times faster than doing things one

[gem5-dev] gem5 gui tool?

Hey folks, a while ago some students developed a GUI tool for gem5. While I think that was great and really hasn't gotten the attention (or adoption?) that a tool like that should, I was looking for it for a more practical reason but haven't been able to find it. Does anybody know where it ended

[gem5-dev] Re: Build failed in Jenkins: compiler-checks #9

Yep, I'll take a look. Gabe On Thu, Oct 14, 2021 at 12:41 PM Bobby Bruce wrote: > Hey Gabe, > > It seems like both the Nightly and Compiler tests are failing because > SPARC no longer compilers. This appears to be due to this change: >

[gem5-dev] Re: reduce the number of checkpoints in the ARM checkpoint test?

ginal Message- > > From: Gabe Black via gem5-dev > > Sent: 12 October 2021 09:18 > > To: gem5 Developer List > > Cc: Gabe Black > > Subject: [gem5-dev] reduce the number of checkpoints in the ARM > > checkpoint test? > > > > I notice that

[gem5-dev] reduce the number of checkpoints in the ARM checkpoint test?

I notice that the long pole in running the quick regressions seems to be the tests/gem5/configs/realview64-simple-atomic-checkpoint.py test which does some sort of ARM linux boot (I assume) and checkpoints 5 times as it comes up. Would it make sense to reduce that down to 3 or even 2? I think that

[gem5-dev] Re: replacement for x86-boot-tests/run_exit.py?

ld function similarly to the old `run_exit.py` but you don't need > to specify the disk and the kernel. > > Kind regards, > Bobby > -- > Dr. Bobby R. Bruce > Room 3050, > Kemper Hall, UC Davis > Davis, > CA, 95616 > > web: https://www.bobbybruce.net > > >

[gem5-dev] flakey kokoro failures

Hi folks, just FYI, I've seen a bunch of kokoro failures recently where there has been only a single line in the log, [00:05:11][ERROR] Config validation failed! https://gem5-review.googlesource.com/c/public/gem5/+/49751

[gem5-dev] replacement for x86-boot-tests/run_exit.py?

Hi folks, particularly Bobby. I was using the x86-boot-tests/run_exit.py configuration to run a fairly quick and easy x86 boot to measure gem5 performance, but that script seems to have gone away in the last week or two. What is the replacement for it? I need something I can use to do a

[gem5-dev] Re: failing ARM dual CPU tests?

? This will guarantee >>> that >>> > you are using the *exact* same "host" setup when running gem5. I think >>> this >>> > is the only way to have a consistent set of tests without something >>> like >>> > Bazel's reproducible build and

[gem5-dev] Re: name collision between tests/gem5 and built in gem5 module

e unless anyone wants to > volunteer) should think about how to clean-up this directory in general. > > -- > Dr. Bobby R. Bruce > Room 3050, > Kemper Hall, UC Davis > Davis, > CA, 95616 > > web: https://www.bobbybruce.net > > > On Wed, Sep 22, 2021 at 10:32 PM Ga

[gem5-dev] name collision between tests/gem5 and built in gem5 module

Hey folks, just FYI, when running tests from the tests directory (as the instructions say to do), if a script imports a gem5 module, there is some ambiguity whether that should be the built in gem5 module added for the components stuff, or if it should be the gem5 directory in the tests directory

[gem5-dev] Re: failing ARM dual CPU tests?

n't work for others as well. Improving the > > documentation so it's more clear how to use the tests or improving the > > testing infrastructure or modifying the tests so that they work more > easily for > > more people would be a *very* welcome contribution. > > > > Cheers, >

[gem5-dev] Re: failing ARM dual CPU tests?

On Wed, Sep 22, 2021 at 9:12 AM Jason Lowe-Power wrote: > Hey Gabe, > > To solve this and the x86 test issue you've been having, I think there are > a couple of possibilities: > > 1. Can you use the docker images that kokoro uses? This will guarantee > that you are using the *exact* same "host"

[gem5-dev] failing ARM dual CPU tests?

Hi folks. When I run the test script main.py locally on an otherwise passing tree, I get 8 test failures. 6 of those are from x86 dynamic linking tests which use a host library which uses a system call gem5 doesn't implement. That is annoying, but I understand that problem. The other 2 are from

[gem5-dev] Re: SCons SimObject() and m5.objects hierarchy

Ack. After thinking about it a bit, I think a minor rename of the riscv version is the right approach for right now. Gabe ___ gem5-dev mailing list -- gem5-dev@gem5.org To unsubscribe send an email to gem5-dev-le...@gem5.org

[gem5-dev] Re: Page size for prefetchers

er. IMO, that's confusing > and error prone. Also, this is the smallest change with the least chance of > unexpected consequences. > > Cheers, > Jason > > On Tue, Sep 14, 2021 at 2:15 AM Gabe Black via gem5-dev > wrote: > >> Hi folks. I'm currently looking at getti

[gem5-dev] Re: SCons SimObject() and m5.objects hierarchy

;> to do that in the future. >> >> Gabe >> >> On Fri, Aug 13, 2021 at 10:21 AM Jason Lowe-Power >> wrote: >> >>> Hey Gabe, >>> >>> Some responses inline below. >>> >>> On Thu, Aug 12, 2021 at 9:37 PM Gabe Black via

[gem5-dev] Page size for prefetchers

Hi folks. I'm currently looking at getting rid of the arch/page_size.hh header file (123 change series for vecregs.hh waiting for review, needs attention from ARM next), and I have at least a good game plan to get rid of the usage of the PageShift and PageBytes constants in most places. The one

[gem5-dev] review help

Hi folks. I've uploaded many reviews recently, and there is another large batch on the way. I really appreciate how there has been good engagement on some of the reviews since I know everyone is busy, and few have a vested interest in my projects. To try to keep the reviews from piling up and

[gem5-dev] SCons refactor reviews

Hey folks, I just uploaded a big new batch of patches to my series which refactors our SCons scripts in various ways and cleans them up. I added reviewers for the first few to get the ball rolling, but to avoid having to go through each one and to avoid a big wave of email to people on the

[gem5-dev] Re: SCons SimObject() and m5.objects hierarchy

at 9:37 PM Gabe Black via gem5-dev > wrote: > >> Hey folks, here are some of my thoughts on the SimObject() mechanism in >> SCons, and the structure of the m5.objects package. >> >> Right now, when you put a SimObject('foo.py') line in a SConscript, that >> ev

[gem5-dev] Re: plan to eliminate the "zero register" concept at the CPU level

Unfortunately it looks like that mechanism I mentioned (for predicating reads/writes) makes some big assumptions which are not generally correct, they just happened to be correct in the situation that feature was initially used. I think I'm going to try to replace that mechanism with one which

[gem5-dev] SCons SimObject() and m5.objects hierarchy

Hey folks, here are some of my thoughts on the SimObject() mechanism in SCons, and the structure of the m5.objects package. Right now, when you put a SimObject('foo.py') line in a SConscript, that eventually leads to SCons importing foo.py (with some special machinery) and discovering through

[gem5-dev] plan to eliminate the "zero register" concept at the CPU level

Hey folks. I was thinking about better ways to handle the "zero register" in the CPUs. A plan I had a little while ago was something sort of like /dev/null, where the zero register would be flattened to one thing if it was a source which would always hold zero, and another thing as a destination

[gem5-dev] Re: overview/documentation/tests for vector register related stuff?

On Mon, Aug 9, 2021 at 1:02 PM Giacomo Travaglini < giacomo.travagl...@arm.com> wrote: > Hi Gabe, > > > -Original Message----- > > From: Gabe Black via gem5-dev > > Sent: 09 August 2021 11:02 > > To: gem5 Developer List > > Cc: Gabe Black > >

[gem5-dev] Re: overview/documentation/tests for vector register related stuff?

I've done a bit of digging so far, and I think I've figured out a bit about the rename mode. 1. This is only used by ARM to handle the difference in how registers are renamed in aarch64 vs otherwise. 2. This is handled in O3 by detecting a squash in the CPU and then checking the aarch64 bit of

[gem5-dev] overview/documentation/tests for vector register related stuff?

Hi folks. I may have asked about this in general before, and I'm pretty sure I've asked about specific pieces of this since I remember Giacomo trying to help me out, but, what resources do we have around: 1. VecRegContainer 2. VecElem 3. VecPredReg 4. rename mode 5. readWriteable accessors 6.

[gem5-dev] Re: kokoro failures due to bug in gem5-art?

;>> On Wed, Jul 21, 2021 at 3:52 PM Bobby Bruce >>>>>> wrote: >>>>>> >>>>>>> There's definitely something funny going on with the gem5art tests >>>>>>> there but I believe that error is happening without trigge

[gem5-dev] Re: python versions used by SCons and gem5

On Thu, Aug 5, 2021 at 3:08 PM Bobby Bruce wrote: > Hey Gabe, > > Just so we're on the same page, I'd like to take a moment to outline how I > understand this problem/solution. Please let me know if I've understood > correctly: > > We essentially have three components of the project which

[gem5-dev] Re: kokoro failures due to bug in gem5-art?

2021 at 3:52 PM Bobby Bruce >>>>> wrote: >>>>> >>>>>> There's definitely something funny going on with the gem5art tests >>>>>> there but I believe that error is happening without triggering a non-zero >>>>>> exi

  1   2   3   4   5   6   >