Practical example for pthread_mutex_trylock

[@muhammetcavus6995]

This series really saved me from watching boring teaching assistant videos. Greatly explained. THANK YOU A LOT.

[@onurcanisler]

*LoL I agree with that:DDD*

[@subbu4u]

Very very good explanation...complete series is awesome

[@narasarajv5278]

Interesting concept for explaining the program...Thank you..

[@katiabibi1966]

very well explained, thank you. But I do not understand why we have "no stove available" in the beginning of execution.. aren't all stoves available at that time?

[@CodeVault]

Ahh it's just the order of printing. Since we have 4 stoves and 10 cooks the first 4 get a stove (but they don't print anything) and the other 6 print out "No stove available", and only after 500ms you'll see the 4 cooks that got a stove print something. You could add the message "Started cooking..." right before: stoveFuel[i] -= fuelNeeded; usleep(500000); And should be clear

[@katiabibi1966]

@@CodeVault thank you so much for the answer !!!

[@probexpd1916]

@@CodeVault Thank you for that answer, Sergui! I keep forgetting that with threads, the sequence of the display doesn't always correlate with what's actually happening. I added "usleep(900000)" like this... usleep(900000); printf("No stove available, wait... "); usleep(300000); Now, the output shows 4 stoves being used, 6 cooks waiting, 4 more stoves being used, 2 cooks waiting and 2 stoves being used _in sequence_

[@mustafa7891]

Nicely explained, it was very helpful : ) But what if a chef occupies a stove that does not have enough fuel, while some other stove have enough fuel for the chef? The chef will still go home, right?

[@CodeVault]

Yea, that's what would happen. You can try and create a mechanism to try other stoves if one is empty of fuel. I kept it simple.

[@prashantpatil3088]

Nicely explain. Please make a video on Shared memory, Message Queue

[@CodeVault]

They are on the to-do list ;)

[@prashantpatil3088]

@@CodeVault thanks bro.

[@atjhamit]

Thanks a lot! Really appreciate the help.

[@maxtirdatov]

Shouldn't i be reset to -1 after sleeping? Otherwise the stove 0 is ignored because i gets incremented before the next iteration.

[@CodeVault]

Ah, you're right, that's a small mistake I made. My bad

[@JamesPetts]

I approve of cooking pots.

[@yahmk3978]

Thank you!

[@iliqnemilov6091]

great video man,thanks a lot

[@denizbatu7812]

excellent

[@beingnikhil155]

Is sleeping for some milli seconds a good practice if all the stoves are busy or we can do something better?

[@CodeVault]

Definitely something better can be done. Like linking it to a condition variable and sleep until someone signals that they stopped using the stove.

[@muhammetcavus6995]

when if(stovefuel[i]-fuelneeded

[@CodeVault]

Yes, you can. Simply unlock the mutex and then continue: pthread_mutex_unlock(&stoveMutex[i]); continue;

[@Codality]

THE BEST

[@Pyrografpl]

Thank you mate :-D

[@HaroldX7677]

Maybe i should be set to -1 so that the loop will start again from 0?

[@CodeVault]

Which of the loops?

[@qrutyq]

Yep it should!

[@qrutyq]

@@CodeVault in the routine one. In other case it will ignore stove with index 0.

[@user-yz7ts2fq9m]

Super!

[@sagivalia5041]

So trylock can be useful if there are more than one critical sections of the same type, right?

[@CodeVault]

Yes, that's one way to use it

[@kippie80]

Would like to see these examples applied to some GPUs.

[@CodeVault]

Using CUDA/OpenCL?