Re: AF_UNIX socketpair dgram queue sizes

2021-11-10 Thread Mark Johnston
On Wed, Nov 10, 2021 at 12:05:33AM -0500, Jan Schaumann via freebsd-net wrote: > Mark Johnston wrote: > > > There is an additional factor: wasted space. When writing data to a > > socket, the kernel buffers that data in mbufs. All mbufs have some > > amount of embedded storage, and the kernel a

Re: AF_UNIX socketpair dgram queue sizes

2021-11-09 Thread Jan Schaumann via freebsd-net
Mark Johnston wrote: > There is an additional factor: wasted space. When writing data to a > socket, the kernel buffers that data in mbufs. All mbufs have some > amount of embedded storage, and the kernel accounts for that storage, > whether or not it's used. With small byte datagrams there ca

Re: AF_UNIX socketpair dgram queue sizes

2021-11-09 Thread Mark Johnston
On Tue, Nov 09, 2021 at 08:57:20PM -0500, Jan Schaumann via freebsd-net wrote: > Hello, > > I'm trying to wrap my head around the buffer sizes > relevant to AF_UNIX/PF_LOCAL dgram socketpairs. > > On a FreeBSD/amd64 13.0 system, creating a socketpair > and simply writing a single byte in a loop t

AF_UNIX socketpair dgram queue sizes

2021-11-09 Thread Jan Schaumann via freebsd-net
Hello, I'm trying to wrap my head around the buffer sizes relevant to AF_UNIX/PF_LOCAL dgram socketpairs. On a FreeBSD/amd64 13.0 system, creating a socketpair and simply writing a single byte in a loop to the non-blocking write end without reading the data, I can perform 64 writes before causing