Category: Committed Code

reapctl() added to DragonFly


Matthew Dillon’s added the reapctl() call, which gives a process control over all sub-processes, even when detached.  This is initally useful for bulk builds of dports, but can probably be extended farther…

Posted by     Categories: Committed Code, DragonFly     1 Comment

Chrome on DragonFly


Chrome runs on DragonFly now, apparently possible now because of this ported fix from Joris Giovannangeli.

Posted by     Categories: Committed Code, DragonFly     1 Comment

DragonFly DRM1 drivers dropped


As Francois Tigeot has pointed out, recent Mesa upgrades have made very old graphics drivers using DRM1 no longer work.  They’ve been removed.  This won’t affect you unless your graphics card is 10+ years old.

Posted by     Categories: Committed Code, DragonFly     1 Comment

/dev/upmap and /dev/kpmap added


Matthew Dillon has added /dev/upmap and /dev/kpmap to DragonFly in two commits.  (er, three.)  It’s an optimization of some sort, and it is unfortunately over my head.  Please, someone comment in a way that fills it in.  I’m tired.

Posted by     Categories: Committed Code, DragonFly     1 Comment

Out of base, into dports


John Marino updated wpa_supplicant (in dports).  He then suggested moving it out of base into dports, so that it could be updated independently of the base system.  (this update, for instance, took years.)   Since wpa_supplicant is necessary to get some systems online – and it can’t be installed if missing if you don’t have a network link – it may be too risky.  I think other packages could be moved out, myself.

Posted by     Categories: Committed Code, DragonFly     6 Comments

OpenSSH update and incompatibility


Robin Hahling volunteered to update OpenSSH in DragonFly, which is good news.  It’s a jump from version 6.1 to 6.7, so there’s some feature changes.  tcpwrappers support is gone, for instance.  If you have a reason to object to this change, speak up now.

Posted by     Categories: Committed Code, DragonFly     1 Comment

Pile of point upgrades


John Marino has upgraded gcc, libedit, xz, and grep in DragonFly.  Also, tzdata has jumped from version 2014e to 2014h, thanks to Sascha Wildner.

Posted by     Categories: Committed Code, DragonFly     0 Comments

UDP improvements


Sepherosa Ziehau’s recent changes to UDP in DragonFly mention some performance gains to sheer packet count.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Lag!


Markus Pfeiffer has imported FreeBSD’s if_lagg to DragonFly.  It’s for talking LACP over multiple network ports, so that the traffic from those multiple ports can be aggregated – if what’s on the other end generally understands LACP.  (Failover mode may not count.)  Please test if you have that sort of surfeit of network ports.

Posted by     Categories: Committed Code, DragonFly, Please test     2 Comments

systat gains -altq display


You can now see the packets, bytes, and drops in altq by using the -altq arg to systat, thanks to this recent commit from Matthew Dillon.

Posted by     Categories: Committed Code, DragonFly     0 Comments

UDP improvements for DragonFly


This very long commit message from Sepherosa Ziehau details the UDP changes he’s made.  It’s mostly technical details, but at the end he mentions this little tidbit:

“For ‘kq_connect_client -u’ test, this commit gives 400% performance improvement (31Kconns/s -> 160Kconns/s).”

Posted by     Categories: Committed Code, DragonFly     0 Comments

An easier moused start


You can now start moused with an argument, so it will look at the right device.  In most cases, I imagine “/etc/rc.d/moused start ums0” will be what anyone wants.  Credit to Michael Neumann for the update.  Perhaps moused_flags="ums0" will do it too?  I haven’t tried yet.

This will overwrite your /etc/devd.conf.

Posted by     Categories: Committed Code, DragonFly     1 Comment

Full rebuild needed


Because of some structure changes made by Matthew Dillon while chasing a pf bug, you will need to do a full buildworld/buildkernel on your next update – if you are running DragonFly-master.  3.8 users are unaffected by the bug or the change.

Posted by     Categories: Committed Code, DragonFly, Heads Up!     0 Comments

New dhclient and other improvements


DragonFly’s dhclient will now retry failed interfaces and handle being re-run gracefully.  This is a blessing for anyone who has had a flaky link.  Matthew Dillon’s made two other improvements for booting that will also improve boot time when networks go missing.

Posted by     Categories: Committed Code, DragonFly     0 Comments

rcreload in DragonFly


Thanks to Zachary Crownover, rcreload is available in DragonFly.  (It’s always good to see a new contributor name.)

Posted by     Categories: Committed Code, DragonFly     0 Comments

libradius updates


Nuno Antunes brought in a significant number of fixes for libradius.  He’s been doing other work recently on netgraph7 support, so I’m linking to this as a ‘signpost’ commit.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Improvements for qemu


While Matthew Dillon was testing the new up-to-256-processor support for DragonFly, he added a few sysctls, one of which helps qemu performance when emulating a lot of processors.  I note it here in case it’s helpful to someone else.

Reserving more memory


DRM (Direct Rendering, not Digital Rights) on DragonFly will normally eat all the memory it thinks it needs.  However, vm.dma_reserved can now be set to a fixed limit in /boot/loader.conf.   By default, vm.dma_reserved on DragonFly is set to 16M, and can be set higher.  I think this is necessary when running higher-resolution screens… Don’t quote me on that, though.

Posted by     Categories: Committed Code, DragonFly     2 Comments

New random algorithms, implementation


Alex Hornung has added a ChaCha algorithms and Fortuna-based CSPRNG to DragonFly’s random device.  You can pick what runs with the sysctl kern.rand_mode, and some other changes.

Posted by     Categories: Committed Code, DragonFly     1 Comment

Default DragonFly/pf keep-policy changed


Matthew Dillon changed the default keep-policy in DragonFly to:

set keep-policy keep state (pickups, sloppy)

This is to match other BSDs (which?  I don’t know) and reduce overhead, according to the commit.

Posted by     Categories: BSD, Committed Code, DragonFly     1 Comment

pf no longer single-threaded


Predrag Punosevac noticed that turning on pf was slowing his machine down.  Rearranging the rules fixed a lot of it for him.  However, Matthew Dillon decided it was time to make pf work concurrently instead of in a single thread, and 24 hours later, it does.  Quick, someone benchmark this!

Posted by     Categories: Committed Code, DragonFly     6 Comments

UDP test, UDP improvements


Sepherosa Ziehau implemented a UDP echo response tool, which not surprisingly meant he also had some UDP performance improvements.  As he points out in the commit, it makes lockless firewall state tables possible.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Random number generator changes


The obvious joke should be “how can you tell?”  Anyway, the csprng in DragonFly has been updated and IBAA is being used more often, and there’s more updates on the way.

Posted by     Categories: Committed Code, DragonFly     0 Comments

GSI CPU Autoselection


Sepherosa Ziehau has enabled GSI target CPU auto selection, by default, on x86_64.  He says to let him know if there’s problems.  I’m not sure what form the problems would take, cause I’m not sure what this does.

Posted by     Categories: Committed Code, DragonFly, Heads Up!     1 Comment

Updates for libpcap, file


libpcap has been updated in DragonFly by Matthew Dillon, and file has been updated by Peter Avalos.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Goodbye, ATM


ATM support is gone in DragonFly, and frankly, I’m surprised it was still there.

Posted by     Categories: Committed Code, DragonFly     0 Comments

ACPICA updated to 20140424


Sascha Wildner’s updated ACPICA to version 20140424.  Will that help you?  Perhaps with newer motherboards; otherwise check the changelog.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Dynamic binaries arrive


Here’s the announcement from Francois Tigeot: DragonFly now uses dynamic binaries in the root filesystem.  You will need to do a full buildworld/buildkernel if on 3.7 and upgrading.

Posted by     Categories: Committed Code, DragonFly, Heads Up!     0 Comments

Rescue initrd added


DragonFly now has a ‘rescue’ system added in, which also functions as a way to mount encrypted filesystems.  Does PAM work yet?  I don’t know; I may be linking to this earlier than I need to.

Posted by     Categories: Committed Code, DragonFly     2 Comments

tcplay updated to 2.0


Alex Hornung has updated tcplay in DragonFly to 2.0, and cryptdisks is updated to match.  If you have a short memory, tcplay(8) is the tool on DragonFly to manage TrueCrypt volumes.  Is DragonFly the only BSD to have this?  I think so, based on very few seconds of googling.

Posted by     Categories: Committed Code, DragonFly     1 Comment

Connection speedup for x86_64


A recent commit from Sepherosa Ziehau has a 5% improvement in the number of network connections per second a x86_64 machine can accept.  He’s also reducing the number of IPIs during network activity.  If this seems somewhat esoteric, it’s because network speeds are getting so fast that the benefits come from reducing the accompanying CPU load.

Posted by     Categories: Committed Code, DragonFly     0 Comments

tzdata2014a and an odd thing


Sascha Wildner updated the time zone database on DragonFly to tzdata2014a.  The odd thing isn’t that update – Sascha updates like clockwork, haha! – but the release notes.  Apparently Even Microsoft is starting to support time zone names, sorta, finally.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Building world for 3.7 users


If you’re on DragonFly 3.7, you will need to build world before building the kernel again if you are updating to some point in the last 24 hours.  Sascha Wildner points out the related commit.

Posted by     Categories: Committed Code, DragonFly, Heads Up!     0 Comments

ACPICA 20140214 brought in


Sascha Wildner brought in ACPICA 20140214, and his commit message has a list of the updates.

Posted by     Categories: Committed Code, DragonFly     3 Comments

Clockmod replaces p4tcc


See the announcement, and the commit.  I’m not totally sure what this affects.

Posted by     Categories: Committed Code, DragonFly     0 Comments

ACPICA-20140114 added


There seems to be a lot of ACPI-related updates lately: Sascha Wildner has updated ACPICA in DragonFly to what I think is the very latest version.  See his commit for the differences.

There really is a daemon in there


John Marino updated daemon(8) on DragonFly.  For some reason, I didn’t know it was a standalone program.  I knew about the idea of daemons as helpers based inside the computer, which is why so many server programs end with a ‘d’ – sshd, ftpd, and so on.  Inexplicably, I never actually saw the program itself.

Posted by     Categories: Committed Code, DragonFly     1 Comment

Other network improvements


Sepherosa Ziehau is continually trying to squeeze more network performance into DragonFly.  I’m not always so good at pointing it out, but here’s several commits from him that improve performance on several chipsets.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Netmap on the way


Franco Fitchner is bringing in netmap to DragonFly.  I don’t think it’s complete yet.

Posted by     Categories: Committed Code, DragonFly     0 Comments

GSoC: SysV IPC code added in


Markus Pfieffer has committed Larisa Grigore’s Google Summer of Code work, “SysV IPC in userspace”.  It’s been a bit since the event finished, but it’s in DragonFly now.

mdocml updated to 1.12.3


Franco Fitchner has updated mdocml in DragonFly to 1.12.3.   The changelog is right on the front page of the vendor site.

Update: Undeadly has a nice summary of the changes.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Discontented with contention? Be content.


Matthew Dillon wrote a roundup post summarizing all the changes he’s made to DragonFly to improve SMP performance in the last few weeks.  He’s removed almost all contention from DragonFly.  This means better performance, scaling upward depending on the number of processors.

‘monster’, the system that builds all 20,000 items in dports, can complete the run in 15 hours.  Compare this to the 2 weeks it used to take me to build the 12,000 packages in pkgsrc.  This is admittedly on different hardware and different packaging systems, but it gives a sense of the scale of the improvement.

 

Posted by     Categories: Committed Code, DPorts, DragonFly, pkgsrc     4 Comments

Possible to poweroff


The ‘poweroff’ command, the equivalent of ‘halt -p’, has been added based on a suggestion from Robin Hahling.

Posted by     Categories: Committed Code, DragonFly     2 Comments

SMP contention improvements


Matthew Dillon was using poudriere, the dports build tool, on a 48-core system.  Poudriere was building all 20,000+ dports, so the machine was quite busy.  He decided to get rid of as much contention as possible, and he’s listed all the ways DragonFly’s been streamlined by these efforts.  We need to revisit some of our previous benchmarks

Posted by     Categories: Committed Code, DragonFly     2 Comments

Continuous dmesg


If you’re watching for a certain event, dmesg(8) on DragonFly now has a -f flag that will display new output as it’s logged, sort of like ‘tail -f’.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Memory statistics changes


While looking for a different bug, Matthew Dillon made some changes in the way memory usage is totaled in DragonFly.  You’ll see this most when using ‘systat -vm 1′ or ‘vmstat’, probably.

Posted by     Categories: Committed Code, DragonFly     0 Comments

gdb, kgdb updated


John Marino has accomplished the major task of updating gdb/kgdb, to version 7.6.1 for DragonFly.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Huge locale changes; full buildworld needed


John Marino has put in a large patch to DragonFly 3.5, updating all sorts of language-related items.  As he warns, you will need a full buildworld/buildkernel in a specific order to update.  On the plus side, you can now probably use your native language for nvi and for git.

Posted by     Categories: Committed Code, DragonFly, Heads Up!     1 Comment

Summer of Code projects getting committed


Matthew Dillon’s committed the work by Daniel Flores on Hammer 2 compression and Mihai Carabas’s vkernel hardware support – both Summer of Code projects.  There’s a good amount of detail in the commit messages describing the work and what it changed; I expect more Summer of Code work to be getting committed…

Note: you’ll want to do a full update.

New systat features


It’s now possible to use systat(1) to see per-connection speeds and pftop status, thanks to Matthew Dillon.

Posted by     Categories: Committed Code, DragonFly     1 Comment

New filesystem for vkernels


Antonio Huete Jimenez has committed his work on “dirfs”, a filesystem that lets you mount directories from your host machine within the running vkernel environment.  It’s a sort of shared folders for vkernels.  See the commit message for usage details.

Posted by     Categories: Committed Code, DragonFly     0 Comments

TTM arrives for DragonFly


Francois Tigeot has ported TTM to DragonFly from FreeBSD and I think a bit from OpenBSD.  All this work has led to an update in the driver porting notes.

Posted by     Categories: Committed Code, DragonFly, FreeBSD, OpenBSD     0 Comments

Just kill everything


killall -T will now kill all processes associated with the current tty, except parents of the killall process itself.  It’s a shortcut to “kill all these runaway items I started by accident”.

Routing stability changes


Apparently Sepherosa Ziehau has been improving DragonFly’s route table performance under extremely heavy load.  (e.g. run efficiently; don’t die)  I don’t have a definitive commit message to point at, but looking at his recent commits are a good start.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Netperf and Jain’s fairness index


Thanks to Sepherosa Ziehau, Netperf will now calculate Jain’s fairness index.  That is a formula that is both interesting and unfamiliar to me.  Not that I understand it, of course – it’s just because it has a neat name.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Recompile needed for DragonFly 3.5 users


Because Sepherosa Ziehau changed mbstat, anyone on bleeding-edge DragonFly will need to rebuild world, or else netstat will become confused.

Posted by     Categories: Committed Code, DragonFly, Heads Up!     1 Comment

New timer available


There’s support for a new timer mechanism in DragonFly 3.5, for x86_64 users: TSC.  Sepherosa Ziehau added support and has described how to disable it – it’s on by default.  It speeds up some very basic (and frequently used) system calls.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Rough network queues added


Sepherosa Ziehau has added a sort of queuing to altq, where TCP ACKs get higher priority.  You may have seen this in any number of pf configurations, where returning data is given its own queue to keep high-volume transfers from slowing themselves down because the acknowledgements can’t get back to the sender.  His commit has statistics on the performance improvement.  He also added a ‘netrate‘ tool for calculating results from using netperf.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Symbol versioning coming in, also buildworld


If you’re using DragonFly 3.5, your next update should be a full buildworld.  That’s because John Marino is adding the framework for symbol versioning.  This means that individual library (.so) files will internally keep track of newer and older symbols.  The current behavior is to name the files differently, which can cause problems if an expected, linked file is missing – even if the needed symbols are present.  The basic framework is being added now, and will be turned on all at once, to minimize the number of times that full buildworld is needed.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Old amd64 removed and extra upgrade step added


The ‘amd64′ specific parts of kernel architecture have been removed, since x86_64 covers all that.  As a side effect of other changes, John Marino warns that upgrading DragonFly from a version older than 3.4, to a version newer than 3.4, will require an intermediate step of going to 3.4 first.  e.g. If your machine is a DragonFly 3.0 system, you will need to upgrade to 3.4 before moving to, say, 3.6 once it is out.  This won’t matter for some months, since the next release is months off.

SO_REUSEPORT turned on


SO_REUSEPORT has been added and turned on by Sepherosa Ziehau.  This is an implementation of a feature found in the Linux kernel.  Check the very lengthy commit message for a description of what it does.

Posted by     Categories: Committed Code, DragonFly     0 Comments

More updates


This time it’s less and libedit, updated by John Marino.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Updates for libmpfd, grep, diff, and libbsdxml


John Marino has committed updates for libmpfr, diff utils, grep, and libexpat/libbsdxml.  Libmpfr, the one item that I suspect doesn’t spring instantly to mind,  is a library for floating-point computation.

Posted by     Categories: Committed Code, DragonFly     0 Comments

ftp (tnftp) update


Peter Avalos has updated ftp in DragonFly.  It’s actually tnftp, which is the same base ftp client used in FreeBSD/NetBSD/Mac OS X/etc.  It’s the 20121224 version, and the 3.4 release branch has it too.

Posted by     Categories: BSD, Committed Code, DragonFly     0 Comments

A new OpenBSD identd


OpenBSD has a new identd daemon.  Is identd used for anything other than verification when connecting to an IRC network?  I’ve never seen it in another context.

Posted by     Categories: Committed Code, OpenBSD     3 Comments

Another sh(1) update


Peter Avalos has committed another batch of updates to sh(1), from FreeBSD.  I was going to comment on how strange it was to see software getting updated so many years later; you’d think everything there was to update for /bin/sh had been done at this point.  Digging casually, the oldest bit on sh that I can find is from 1991 – 22 years old.   The man page mentions a rewrite in 1989 based on System V Release 4 UNIX, and there were versions of sh all the way back to version 1.

Here’s a trivia question – what’s the oldest Unix utility, and what’s the oldest code still in use?  I don’t know the answer.

Posted by     Categories: BSD, Committed Code, DragonFly, FreeBSD, UNIXish     4 Comments

3 very different commits


Here’s 3 recent and different commits to DragonFly that I’m commenting on all at once:

  1. Peter Avalos upgraded libarchive in DragonFly to 3.1.2, with a note of the changes.  An ordinary and appreciated update.
  2. Sascha Wildner updated the ISO639 file to include the newest update: “Standard Moroccan Tamazight”.  There’s no particular utility to that; I just like saying “Standard Moroccan Tamazight” out loud.
  3. Work on poudriere, the utility for bulk-building DPorts packages, has caused some nice speedups for DragonFly in extremely stressful situations.  See one of Matthew Dillon’s recent commits.

I really wish the other BSD projects would include commit lines in the mail message subjects, so it was easier to catch things like these.

Posted by     Categories: BSD, Committed Code, DPorts, DragonFly     0 Comments

Read shortcut, buffer cache improvements


The new vm.read_shortcut option has been turned on by default by Matthew Dillon, which should lead to some performance improvements.  That improvement has been measured for tmpfs, at least.  There’s also some buffer cache improvments that help on x86_64 systems, too.

Update: As Venkatesh Srinivas pointed out, tmpfs also no longer uses the mplock, so it’ll take better advantage of multiple processors.

Posted by     Categories: Committed Code, DragonFly, Goings-on     0 Comments

World backups when upgrading


John Marino proposed a method for backing up world when upgrading, for those rare but catastrophic cases where the installed programs can’t run.  After some discussion, he committed an automatic backup method, and there’s a ‘restoreworld’ target to take advantage of it.

The kernel already gets renamed to kernel.old as a backup, if I remember correctly.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Per-CPU network stats


As Sepherosa Ziehau mentions in his latest commit, DragonFly now collects IFNET/IFADDR statistics on a per-CPU basis.  This makes it more accurate, but may mess with any third-party program that accessed it directly.  I don’t know if there’s anything in pkgsrc that does that…

Posted by     Categories: Committed Code, DragonFly     0 Comments

Updates: OpenSSL, libdialog, tzsetup, locale


I know OpenSSL in DragonFly was just updated, but Peter Avalos has done it again, bringing it to version 1.01e.  I assume this new version is to fix some recently-exposed problems.   He also has updated libdialog, which was previously not located in contrib/, as sime third-party software needed a more modern version.  As a side effect from that, tzsetup in DragonFly now matches the version in FreeBSD and NetBSD.  And, Sascha Wildner has updated the locale files on DragonFly, also to match FreeBSD and NetBSD.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Comings and goings


Added: Peter Avalos has updated OpenSSL to version 1.0.1d – see the changelog.

Removed: support for ISA sound cards, by Sascha Wildner.  Goodbye sb16; I’ll remember you fondly.

Posted by     Categories: Committed Code, Device support, DragonFly     0 Comments

Multiple TX queue support for emx(4), sort of


The emx(4) driver now has support for multiple TX queues, but it’s not on by default.  There’s scenarios where multiple queues work out with that hardware, but you have to be sure you are actually in the right setup for that first.  Check Sepherosa Ziehau’s commit message for the details.

Posted by     Categories: Committed Code, Device support, DragonFly     0 Comments

GCC 4.7 the new default


John Marino has set gcc 4.7 as the default compiler in DragonFly.  This replaces the previous default of gcc 4.4.  The 4.4 version is still available, and while you can set NO_GCC44 to keep it from being built, John’s commit message notes that it’s still useful especially for some ports that don’t work with gcc 4.7.

Posted by     Categories: Committed Code, DragonFly     0 Comments

GCC 4.7 for DragonFly 3.4?


GCC version 4.7 is already available now in DragonFly 3.2, but it’s not the default compiler.  John Marino intends to make it default for the next release.  What’s that mean for us?  Nothing other than a new compiler, since he’s already fixing related issues.

Posted by     Categories: Committed Code, DragonFly     0 Comments

Network fairness changes and what they mean


Sepherosa Ziehau makes commits almost daily to DragonFly’s network infrastructure, but I have a hard time quantifying it into Digest posts in part because it’s often very technical.  His most recent commits come with an explanation, however.  He has done plenty of work to improve overall transmission speeds in DragonFly, and now he’s working on ‘fairness’.  Fair, in this case, means ensuring that packet transmitting and receiving happen without either one monopolizing the connection.  In real world terms, this translates to much more constant speeds.  His recent commit details what he’s doing and some numbers to prove it.

Remember I said he’s improved speeds?  Note that in his example, he’s reaching stable peaks of 981 Mbps.  This is on a line that I assume theoretically maxes out at 1000.

Posted by     Categories: Committed Code, Device support, DragonFly     0 Comments

Updates for m4 and flex


Peter Avalos has updated m4 for DragonFly.  This will bring us a little more in sync with the other BSDs.  Also, John Marino has updated flex, which is apparently 17 years old?   Meaning it hasn’t been updated in DragonFly ever, and then not in FreeBSD before that, for a long time.  Looking at the timeline on the flex web page appears to match.

Posted by     Categories: BSD, Committed Code, DragonFly     0 Comments

Multiple TX queue support started


Sepherosa Ziehau has added a generic form of support for multiple transmit queues in DragonFly.  This means less contention when transmitting.  It’s not done; he has drivers to set up and as he said, it’s “step 1 of many“.

Posted by     Categories: Committed Code, DragonFly     0 Comments

3.3 users, please update


If you are on DragonFly 3.3, and you are running a kernel built after January 1st, there’s a bug in the way FP context is handled when the kernel supports AVX.  (January 1st is when AVX support was committed.)  Matthew Dillon has committed a fix and issued a note to update for everyone.

Posted by     Categories: Committed Code, DragonFly, Heads Up!     0 Comments

More on the Himeno Phoronix benchmark, and memory allocation


If you recall, Phoronix recently ran a bunch of benchmarks on DragonFly.  One spot that didn’t look good was the “Himeno Poisson Pressure Solver”.  I’m no closer to knowing what capability it actually tests other than itself, but Alex Hornung, Matt Dillon, and Venkatesh Srinivas figured out that cache coloring was the missing ingredient.  DragonFly now scores the same as Linux.

Tangentially related, this cache coloring is happening in nmalloc, which is now used on 64-bit DragonFly systems.  The previous one, dmalloc, had problems in long-running programs.

Posted by     Categories: Committed Code, DragonFly     0 Comments