15 Dec
2022
15 Dec
'22
11:36 a.m.
Don't see how unless they put multiple related services in the same
address space, which reduces context switching but tends toward a
monokernel (& increased coupling). Unless I am misunderstanding you.
On Dec 14, 2022, at 6:54 PM, Larry McVoy
<lm(a)mcvoy.com> wrote:
Wasn't there some statement that QNX dropped some of these? Copy plus
context switch?
On Wed, Dec 14, 2022 at 04:29:45PM -0800, Bakul Shah wrote:
On Dec 11, 2022, at 7:09 PM, Andrew Warkentin
<andreww591(a)gmail.com> wrote:
--
---
Larry McVoy Retired to fishing http://www.mcvoy.com/lm/boat
It's not necessarily true that microkernels are significantly slower.
uKernels are usually quite fast as they do so little. What can be slow
is emulating a Unix like OS on top due to context switches. For instance,
a user process doing read() will have the following context switches:
userProc->uK->FileSystem->uK->diskDriver->uk->FileSysem->uK->userProc
or worse (I didn't account for a few things). Because of this even some
uKernels run a few critical services + drivers in the supervisor mode.
But overall slowdown of such a unix emulation will very much depend on the
workload and also what kind of performance improvements you are willing to
try in a complex kernel vs same services running in user mode.
At present the linux kernel has about 31+ Million lines (accounting for
all architectures, filesystems, device drivers etc.). The FreeBSD 13.x
kernel is about 8.7M LoC (of which 44-45% are in device drivers). I only
counted .c and .h files. In contract FreeBSD 2.2.2 kernel has ~554K LoC.
This LoC growth is entirely understandable but I wonder how things may
have turned out in an alternate universe of uKernel based designs....