Timeline for Main stacks in Linux
Current License: CC BY-SA 3.0
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 23, 2016 at 7:21 | comment | added | kaiwan | AFAIK, nested interrupts of different IRQ lines are allowed. NOT the same IRQ! IOW, when handling interrupt 'n', interrupt 'n' will be turned Off, i.e., it is non-reentrant. Of course, top and bottom halves can still execute concurrently.. | |
| Dec 30, 2015 at 11:25 | comment | added | Nik Novák | @Netch So I asked about that. Here is an link: stackoverflow.com/questions/34527763/… | |
| Dec 30, 2015 at 10:04 | comment | added | Nik Novák | @Netch This sounds like a powerfull solution, BUT. I am afraid about that Linux uses nested interrupts... | |
| Dec 30, 2015 at 8:37 | comment | added | Netch |
Handling of another external interrupt during a running one is not allowed, AFAIK, in Linux. This could be too resrictive for some cases but greatly simplifies design. Multiple interrupt handlers in parallel are allowed in SysV and BSD design (look for splbio, splnet, etc.). Anyway, interrupt handlers are designed in manner so they limit only to parts that can't be performed in softinterrupt state. There are "realtime" versions which allow multiple interrupt handlers; see my answer for generic pecularities which occur with such permission.
|
|
| Dec 29, 2015 at 22:54 | history | edited | vonbrand | CC BY-SA 3.0 |
"Stack", not "kernel"
|
| Dec 29, 2015 at 21:58 | comment | added | Nik Novák | Thank you for your answer. I can imagine the situation when an interrupt occurs and inside its handler enables further interrupt (nesting interrupts). So, what if stack size is small? What is occurs overflow? | |
| Dec 29, 2015 at 21:21 | history | answered | vonbrand | CC BY-SA 3.0 |