Segmentation fault (core dumped)

>Segmentation fault (core dumped)

Other urls found in this thread:

reddit.com/r/Amd/comments/7u6sld/ryzen_segfault_bug_in_a_postweek25_cpu/
twitter.com/SFWRedditVideos

also what does core dumped mean

>accessing memory past /0

At least you got debug. Days numbered folks.

a file was written to the app's last known working directory (usually the location from where you launched it)
[despair@bsdnox src]$ some_app
.
.
.
Segmentation fault (core dumped)
[despair@bsdnox src]$ ls
-rw-r--r-- 1 despair wheel 131072054 Jan 1 19:49 some_app.core

It contains just enough memory context for a dev to load the dump file plus the program that created it, inside a debugger to find the cause of the problem after-the-fact.

when do seg faults occur?

when you read or write memory you're not allowed to

Core dumped comes from the old days where memory on computers were cores made from a ferrous material.

OS gives your program some pages of memory it can access. It you try to access data outside of those permitted, the MMU detects that and sends interrupt to processor which then tells the OS you are a nasty program and should be punished.

Apparently, some Ryzen processors manufactured in week 25 (and some later on, though drastically less occuring) suffer from segmentation fault bug. It appears to be happening the majority of the time in Linux. One easy way to find out is to run the "kill ryzen" script from Phoronix

reddit.com/r/Amd/comments/7u6sld/ryzen_segfault_bug_in_a_postweek25_cpu/

punish me daddy

>OS doesn't have segfaults
>computer catches fire

When a programming accesses memory locations that is not assigned to it
Ex.
int* hello = malloc(sizeof(into));//ask for an address that is the size of int
*hello=8;//assign 8 to that address
*(hello-sizeof(Int))=8;//assign 8 to the address before that , that address is not assigned specifically to that program and thus it ((can)) triger a segfault

i wish the worst thing in programming was segfaults

When a programming accesses memory locations that is not assigned to it
Ex.
int* hello = malloc(sizeof(int));//ask for an address that is the size of int
*hello=8;//assign 8 to that address
*(hello-sizeof(Int))=8;//assign 8 to the address before that , that address is not assigned specifically to that program and thus it ((can)) triger a segfault

Guru Meditation

Sounds a lot more positive than "fault"

This isn't completely true, computers use virtual memory so it would just be an unallocated virtual address, and malloc will actually allocate more than you request, generally it will give you a full page, so it's possible that the memory before your variable is actually fully writable

>mummy overflowed my buffer.

I had Firefox seg fault when I filled up RAM with tmpfs

>mum found the stacktrace

>In Linux
>lintoddlers bought a windows gaming chip
And nothing of value was lost

Well when an exception occurs, the kernel will then look for exception handlers in the process to hand it off to, shit like vectored exception handlers, structured exception handlers, etc
It goes through each one of the handlers> , passing them information about the exception, I.e. the exception that occurred, the state of the registers, where the exception occurred
If one of the exception handlers is set up to catch that specific exception, it will shout hey it's me I can fix this, will do it's best to fix the exception and then continue execution at where the exception occurred
If the handler isn't able to fix the exception, it will hand it off to the next in the list, until it reaches either one that can handle the exception or the unhandledexceptionfilter
If it hits the unhandledexceptionfilter, the process will create a mini dump of itself and write it to disk, allowing software engineers to view the call stack/registers for when that exception occurred, and fix whatever bug caused it

kill -n SIGSEGV $(pidof $0)

Wat

How did you solve the captcha?

He bought a pass lol.

from testing it didn't always segfault , now i know why thanks

all bash instances will close right ?

Are you stupid?
Captcha works fine on iPhone :)

>implying the 8086 or 8088 were housefire-tier
>implying that any CPU simple enough to lack memory protection was housefire-tier

>vectored exception handlers
this is an 18+ site