Why is setscheduler not forcing need resched
- by EpsilonVector
I noticed that in Linux kernel 2.4 setscheduler doesn't force need_resched. Why is that? is it just some convention, or does that happens somewhere else?
Search Results
Search found 56 results on 3 pages for 'epsilonvector'.
Page 3/3 | < Previous Page | 1 2 3
-
-
Context switches in control paths of a non-preemptive kernel (Linux)
- by EpsilonVector
The Linux kernel is non-preemptive, but I just read that there could be context-switches in different control paths. Doesn't that contradict the non-preemptive nature on the Linux kernel?Read the article
-
Preventing threads from writing to the same file
- by EpsilonVector
I'm implementing an FTP-like protocol in Linux kernel 2.4 (homework), and I was under the impression that if a file is open for writing any subsequent attempt to open it by another thread should fail, until I actually tried it and discovered it goes through. How do I prevent this from happening? PS: I'm using open() to open the file.Read the article
-
Write failed, errno 0
- by EpsilonVector
I have a client server situation in which I receive data using read(socket, char_buf, BUF_SIZE) and then try to write it into a log file using write(filefd, char_buf, strlen(char_buf)) Strangely enough this fails (write returns -1), and yet the errno is set to 0, and I can print the message, AND the log file descriptor works (I write to it before and after this command). What's going on?? (Working on Linux kernel 2.4 (homework))Read the article
-
What is the best way to identify that a function returned and automatically do something with it?
- by EpsilonVector
I'm implementing user threads on a Linux 2.4 kernel (homework) and my threads are basically just functions running with their own execution context (for example: uthread_create(functionpointer, args)). I'm having a hard time figuring out how to catch the return value and have it ready to be returned when two threads join. I know where to get the value from (eax), but not how to identify that the function actually finished. There's something called atexit, but I've seen people advising against using it, so... what is the best way to discover that a function returned and do something about it?Read the article
-
How to explain a layperson why a developer should not be interrupted while neck-deep in coding?
- by András Szepesházi
If you just consider the second part of my question, "Why a developer should not be interrupted while neck-deep in coding", that has been discussed a number of times by smart people. Heck, even the co-founder of SO, Joel Spolsky, wrote a blog post about "getting in the zone" and "being knocked out of the zone" and why it takes an average of 15 minutes to achieve productivity when participating in complex, software development related tasks. So I think the why has been established. What I'm interested in is how to explain all that to somebody who doesn't know beans about Beans (khmm I mean software development). How to tell the wife, or the funny guy from accounting at the workplace, or the long time friend who pings you on Skype every 30 minutes with a "Wazzzzzzup?!", that all the interruptions have a much deeper impact on your work than the obvious 30 seconds they took from your time. Obviously you can't explain it by sentences like "I have to juggle a lot of variable names in my short term memory" unless you want to be the target of blank stares or friendly abuse. I'd like to be able to explain all that to non-developers in a way that will make them clearly understand - without being offensive, elitist or too technical. EDIT: Thanks to everyone for their great insights. I've accepted EpsilonVector's answer as his analogy was the closest one to my original needs. The "falling asleep" explanation is neither offensive nor technical, almost anyone can relate to it, and the consequences of getting disturbed while falling asleep or while being in the zone are very similar: you experience frustration and you "lose" 15-20 minutes of time.Read the article
