Cover photo
Verified name
Linus Torvalds
Works at Linux Foundation
Attended University of Helsinki
Lives in Portland, OR


Linus Torvalds

Shared publicly  - 
Woot! Another release of Subsurface approaches.

Of course, I'm also doing the 3.16-rc5 release of the kernel today, but with kernel RC's being a regular weekly thing, I've grown bored with even mentioning them. So I mention +Dirk Hohndel's releases instead.
Subsurface 4.2 BETA1 is out. Test it while it's fresh...
Jim Carver's profile photoCheco IMG's profile photoMike Murphree's profile photoKrzysztof Wierzbicki's profile photo
Tell +Dirk Hohndel to stop fooling around with the quad copter and focus!
Add a comment...

Linus Torvalds

Shared publicly  - 
Yo Divers! You know the drill. Prettier, more capable, and fewer bugs.
Subsurface 4.1 has been released

Tons of bug fixes, a very nice visual update, lots of improvements overall.
Windows and Mac binaries are on our server, sources are available via git or as tar ball on our server. Packages for the different Linux distributions will show up over the next few days / weeks / months (depending on the nature of your favorite distro... but realistically, just build it yourself...)
Valent Turkovic's profile photoCri Ort's profile photoRoy Andre Tollefsen's profile photoJoe Kremer's profile photo
who r u?
Add a comment...

Linus Torvalds

Shared publicly  - 
Hey divers! Subsurface 4.1 is just around the corner. 

Fancier dive profiles (Animated when switching dives! More accessible profile settings!). And you can save to git branches.
And we all know what comes after Beta 1, right?
Oh well, you had to be there, it really makes sense... happy to announce Beta 3...
Go out and test, please
Finjon Kiang's profile photoHoward Pepper's profile photoBen Roberts's profile photoTed Zhai's profile photo
Andrew Semborski - sick with type 1 diabetes, among other things, has a chronic illness such as heart failure, heart attack which threatens him, diabetic polyneuropathy, hiatal hernia, spinal degeneration, and sciatica, which caused partial paralysis of his right leg ...

Therefore need immediate help to be able to buy it needed insulin, and medications, lack of insulin threatens his health and even life. Mr. Andrew does not have any financial resources. In view of the above, please All people of good heart for any help! ..
Add a comment...

Linus Torvalds

Shared publicly  - 
One of the things I end up doing is do a lot of performance profiling on core kernel code, particularly the VM and filesystem. 

And I tend to do it for the "good case" - when things are pretty much perfectly cached.  Because while I do care about IO, the loads I personally run tend to be things that cache well. For example, one of my main loads tends to be to do a full kernel build after most of the pulls I do, and it matters deeply to me how long that takes, because I don't want to do another pull until I've verified that the first one passes that basic sanity test.

Now, the kernel build system is actually pretty smart, so for a lot of driver and architecture pulls that didn't change some core header file, that "recompile the whole kernel" doesn't actually do a lot of building: most of what it does is check "ok, that file and the headers it depends on hasn't changed, so nothing to do". 

But it does that for thousands of header files, and tens of thousands of C files, so it all does take a while. Even a fully built kernel ("allmodconfig", so a pretty full build) takes about half a minute on my normal desktop to say "I'm done, that pull changed nothing I could compile".

Ok, so half a minute for an allmodconfig build isn't really all that much, but it's long enough that I end up waiting for it before I can do the next pull, and short enough that I can't just go take a coffee break.

Annoying, in other words.

So I profile that sh*t to death, and while about half of it is just "make" being slow, this is actually one of the few very kernel-intensive loads I see, because it's doing a lot of pathname lookups and does a fair amount of small short-lived processes (small shell scripts, "make" just doing fork/exit, etc).

The main issue used to be the VFS pathname lookup, and that's still a big deal, but it's no longer the single most noticeable one.

Most noticeable single cost? Page fault handling by the CPU.

And I really mean that "by the CPU" part. The kernel VM does really well. It's literally the cost of the page fault itself, and (to a smaller degree) the cost of the "iret" returning from the page fault.

I wrote a small test-program to pinpoint this more exactly, and it's interesting. On my Haswell CPU, the cost of a single page fault seems to be about 715 cycles. The "iret" to return is 330 cycles. So just the page fault and return is about 1050 cycles. That cost might be off by some small amount, but it's close. On another test case, I got a number that was in the 1150 cycle range, but that had more noise, so 1050 seems to be the minimum cost.

Why is that interesting? It's interesting, because the kernel software overhead for looking up the page and putting it into the page tables is actually much lower. In my worst-case situation (admittedly a pretty made up case where we just end up mapping the fixed zero-page), those 1050 cycles is actually 80.7% of all the CPU time. That's the extreme case where neither kernel nor user space does much anything else that fault pages, but on my actual kernel build, it's still 5% of all CPU time.

On an older 32-bit Core Duo, my test program says that the page fault overhead is "just" 58% instead of 80%, and it does seem to be because page faults have gotten slower (the cost on Core Duo seems to be "just" 700 + 240 cycles).

Another part of it is probably because Haswell is better at normal code (so the fault overhead is relatively more noticeable), but it was sad to see how this cost is going in the wrong direction.

I'm talking to some Intel engineers, trying to see if this can be improved. 
Sikhwari Phathutshedzo's profile photoBret Towe's profile photoMaria Aparecida's profile photoTomas Hruby's profile photo
Add a comment...

Linus Torvalds

Shared publicly  - 

When you do this the first time, it looks like a simple oversight - left-overs from some debugging perhaps. But when you "fix" it by hiding it, it's very clearly deliberate.

Netgear, Cisco, quick question please? What exactly does it take to actually care about your customers instead of actively trying to screw them?
Researcher finds secret “knock” opens admin for some Linksys, Netgear routers.
Abdulhalim Kuchusev's profile photoEd Daniel's profile photoRudolf E. Steiner's profile photoJono Knudsen's profile photo
+Dan Kolář one's feature is someone else's bug
Add a comment...

Linus Torvalds

Shared publicly  - 
So last year I was at the Women of Vision awards gala in the Bay Area. Almost exactly a year later, I attended a rather smaller, more local but more personal version of that: NCWIT Award for Aspirations in Computing.

Congrats to Patricia.
Sedat Kapanoglu's profile photoDirk Hohndel's profile photoLydia Stench's profile photoJoshua Cavell's profile photo
Congratulations to Patricia! 
Add a comment...

Linus Torvalds

Shared publicly  - 
Congrats to Patricia.

As to everybody else: you might want to avoid Oregon (and nearby states) for the next few months. It's beautiful here in the summer, but you have to ask yourself: do you feel lucky? 
Stuart McMurray's profile photoRani Ahmad's profile photoOleksandr Natalenko's profile photoAnna Casiero's profile photo
Regarding drinking in general it should be banned at least ,if not a complete ban, outside your homes. At home you're free to do whatever you like. No drinking means that any country can save allot of money that would be spent on others in need , maybe cancer, instead of wasting it on people who are technically drinking fuel. Alcohol is really car fuel!!! I just don't get it how you guys so highly educated and defend drinking as a whole. Wanna have fun??? There are heaps of ways to have fun. Wanna forget??!!?? Yes! I now found the way to forget without drinking: just think positive! That's all you need to do. 
Add a comment...

Linus Torvalds

Shared publicly  - 
[ Insert "I need you" propaganda picture here ]

Are you familiar with the OS X bluetooth stack, and how to use it from C? Do you crave fame and fortune and an open source project to show off your mad OSX BT skillz in? 

If so, you're in luck. The libdivecomputer project needs some help with the OS X bluetooth implementation, because apparently it is completely different from the Linux/Windows ones that already has some preliminary patches.

See the link for details.
François Varas-Genestier's profile photoRi Snyman's profile photoDirk Hohndel's profile photoElena Kuprienko's profile photo
Add a comment...

Linus Torvalds

Shared publicly  - 
Hey, +Gmail, your spam filers seem to be having trouble.

I started looking through my spambox because Al Viro reported that he got spam-warning bounces when emailing me. His email actually arrived, and nothing from him was in the spam traps, but something certainly seems to have changed since April 23rd: there was a definite uptick in false positives in my spam folder.

Maybe it's just a coincidence that there seems to be an uptick in wrongly marked spam in the last week, but have you perhaps been tweaking gmail spam detection lately? Because if you have, I think it might have become a bit too aggressive.

Of course, maybe it's just that there does also seem to be more spam senders that use Markov chains to make their messages look more like real email (or just cut and paste real email into the bottom of the spam message). So maybe that's poisoning your Bayesian filters.
Vinay Paul's profile photoAleksandr Bely's profile photoDavid Stevens's profile photoGameOver RIP's profile photo
I found few of my voice invoices in spam from +Flipkart. And my friend found his apple developer license in spam. Don't know what's wrong with spam filters.
Add a comment...

Linus Torvalds

Shared publicly  - 
Ze Frank's treatise on the coolest of creatures.
I'm thrilled every time I see an octopus. They are so amazingly cool.
Daniel Lench's profile photoKenneth Leisti's profile photoChris Ludlow's profile photoScott Lantow's profile photo
Add a comment...

Linus Torvalds

Shared publicly  - 
We're apparently the only people around who have random videos of our kids doing various things shot with random phones or cameras that don't always get the rotation right. 

Because I just went through several different video players until I found one that allowed me to rotate the video to be right side up.

smplayer for the win.

But that's after trying too many alternatives.

Even VLC, that does pretty much anything and has an overly fancy video geometry thing that allows you to rotate the video by an arbitrary amount gets it completely wrong. Not only does it make it unnecessarily hard to actually do a simple 90° rotation (because obviously you want an analog input thing that makes it possible to rotate by something like 82°, but no obvious way to just fix a vertical video?), but then it keeps the end result within the original aspect ratio box so that the skewed end result is wrongly cropped. I'm sure you can use another setting to edit that aspect ratio too, but after that amount of crazy who cares any more?

And the fact that we must be the only ones to ever encounter this is odd, since google actually finds a lot of people asking for this (usually with the answer being "use ffmpeg to re-encode the whole video correctly", which is one of those answers that are "technically correct, but incredibly stupid").

I'm sure there are plugins for other players, but I would like to congratulate the smplayer developers for just making it easy for people. And it even seems to remember the setting (per video!) so that once you've corrected one video you don't have to correct it over and over again.
Gustavo Solís's profile photoAntonio Bucciol's profile photoAff Vaf's profile photoFernando Miguel's profile photo
Hola Linus Soy una joven secretaria de origen Peruana. I speak english more an less.  I prefer spanish.  Do you speak spanish?.  bye
 ·  Translate
Add a comment...

Linus Torvalds

Shared publicly  - 
Obey the penguin!
Judging by the size of linux-next for version 3.15, the next release is going to be one of the bigger releases (in number of commits) in some time.

And while the merge window has another week to go, what is in the current -git tree should be the bulk of the merges. There will certainly be more, but we've already cracked the 10k commit limit.

So if you are feeling adventurous, and don't mind compiling a kernel out of a git tree, and like the notion of starting testing early, I must encourage you to try now. The sooner we make sure we've dotted all the i's and crossed all the t's, the better off we'll be.

Get it first, get it fresh:
Mike Grunewald's profile photoOzcan Atacan's profile photoMutz Metz's profile photoBitter Schnapps's profile photo
Dear Mr Torvalds I wonder if you can guide me.  I am a relatively new user to linux.  I like to have choice and to this  end I use and play with different distros which include Arch, Aptosid and Gentoo.. I have noticed that there is a drive towards "systemd" and irrespective of users opinion this is implemented (in the case of Arch Linux). Now I have learn't my way around linux in the true hacker spirit, by practice, learning from other people and by my own mistakes.  However there are certain people that I follow (who are competent linux users) like "ignorant guru" and others who are not happy with the switchover to the point that they have pulled out of developing for it. Please can you address this problem  .. because it is becoming one..  I personally have decided to switch over to Gentoo because they give the "option" to use OpenRC.. the following link is for ignorantguru take on the implementation of "systemd"  -   I really would like you opinion.. thank you in advance..
Add a comment...
Creator of Linux and git
  • University of Helsinki
Basic Information
SW Engineer
  • Linux Foundation
    SW Engineer, present
  • OSDL
  • Transmeta Corp
Map of the places this user has livedMap of the places this user has livedMap of the places this user has lived
Portland, OR
Helsinki, Finland - Bay Area, CA
I've been to Fish 'n fins twice now, once during off-season (June), and once during high season (November). Very professional. They clearly get very crowded during high season, but seem to successfully juggle having multiple boats out. Good diving, good people.
Public - 7 months ago
reviewed 7 months ago
1 review