Implement mechanism to wait on any of several futexes • Hacker News Board

Re: Implement mechanism to wait on any of several futexes

#11

Post by kleer001 » Wed, Jul 31, 2019, 4:15 PM UTC

sjwright wrote: post #10

belltaco wrote: post #7

Earlier quoted context omitted.

4.11 is 9 versions newer than 4.2

The thing to remember is that version numbers aren’t decimals. They’re two or three integers separated by periods. 11 is higher than 2.

Harumph. Then it should be padded with zeros. So, 4.002 and 4.011

/grumpy

edit: nope, it's fine as is. my experience in frame numbers, dates, and so forth do not scale to releases

Re: Implement mechanism to wait on any of several futexes

#12

Post by msaadat » Wed, Jul 31, 2019, 4:15 PM UTC

scohesc wrote: post #5

Am I interpreting the graph on their page all wrong? It looks like the older version of Proton provides better performance? Unless they're graphing function call return time? The graph and the text around it doesn't do a good job of explaining that.

Graph seems to be FPS

Re: Implement mechanism to wait on any of several futexes

#13

Post by Kuraj » Wed, Jul 31, 2019, 4:17 PM UTC

aardvarklegend wrote: post #8

Kuraj wrote: post #3

The scale goes from 0 to 80. It's not horribly misleading, but just saying

I believe that’s because it’s frames per second, not an abstract percentile.

Oh, wow.

I'm an idiot

Re: Implement mechanism to wait on any of several futexes

#14

Post by rimunroe » Wed, Jul 31, 2019, 4:17 PM UTC

sjwright wrote: post #10

belltaco wrote: post #7

Earlier quoted context omitted.

4.11 is 9 versions newer than 4.2

The thing to remember is that version numbers aren’t decimals. They’re two or three integers separated by periods. 11 is higher than 2.

I don't know what the source of this convention is, but this is why I was taught that the version "1.2.3" should be read aloud as "one dot two dot three" instead of "one point two point three". The idea is that people--where I'm from at least--tend to read decimals as "point" and not "dot".

Re: Implement mechanism to wait on any of several futexes

#15

Post by rbanffy » Wed, Jul 31, 2019, 4:22 PM UTC

sedatk wrote: post #4

This is apparently for improving the performance of Wine emulator on Linux mainly, not native Linux games like the headline suggests.

Would it be useful to native programs?

Re: Implement mechanism to wait on any of several futexes

#16

Post by Sahhaese » Wed, Jul 31, 2019, 4:22 PM UTC

robin_reala wrote: post #2

The specific proposal (linked to from the article) is https://lkml.org/lkml/2019/7/30/1399

Depressing to see reviewers waste review bandwidth bringing up issues such as "wasted newline" and "incorrect comment format". Do kernel developers not use auto-formatters?

Re: Implement mechanism to wait on any of several futexes

#17

Post by Someone1234 » Wed, Jul 31, 2019, 4:24 PM UTC

rbanffy wrote: post #15

sedatk wrote: post #4

This is apparently for improving the performance of Wine emulator on Linux mainly, not native Linux games like the headline suggests.

Would it be useful to native programs?

Wine is a native program, so yes. But it is a niche facility that Wine benefits from, and other programs that handle a lot of locks/queues also do, but may not directly benefit Linux native game performance.

Re: Implement mechanism to wait on any of several futexes

#18

Post by golergka » Wed, Jul 31, 2019, 4:24 PM UTC

kleer001 wrote: post #11

sjwright wrote: post #10

Earlier quoted context omitted.

The thing to remember is that version numbers aren’t decimals. They’re two or three integers separated by periods. 11 is higher than 2.

Harumph. Then it should be padded with zeros. So, 4.002 and 4.011 /grumpy edit: nope, it's fine as is. my experience in frame numbers, dates, and so forth do not scale to releases

But why? You don't know how many decimals you'll need in advance - three digits isn't enough for Windows build number, for example.

Re: Implement mechanism to wait on any of several futexes

#19

Post by kinleyd » Wed, Jul 31, 2019, 4:25 PM UTC

kleer001 wrote: post #11

sjwright wrote: post #10

Earlier quoted context omitted.

The thing to remember is that version numbers aren’t decimals. They’re two or three integers separated by periods. 11 is higher than 2.

Harumph. Then it should be padded with zeros. So, 4.002 and 4.011 /grumpy edit: nope, it's fine as is. my experience in frame numbers, dates, and so forth do not scale to releases

Some years ago I thought so too. Now I'm just used to it.

Re: Implement mechanism to wait on any of several futexes

#20

Post by ufo » Wed, Jul 31, 2019, 4:26 PM UTC

rbanffy wrote: post #15

sedatk wrote: post #4

This is apparently for improving the performance of Wine emulator on Linux mainly, not native Linux games like the headline suggests.

Would it be useful to native programs?

Valve's post suggests that they believe this to be the case, although they didn't explain the specific details.

> We think that if this feature (or an equivalent) was adopted upstream, we would achieve efficiency gains by adopting it in native massively-threaded applications such as Steam and the Source 2 engine.