Re: [Qemu-devel] [PATCH 4/5] uhci: rewrite UHCI emulator, fully async operation with multiple outstanding transactions

[Max Krasnyansky]

Hi Juergen,

Juergen Lock wrote:
> This is whats got committed as r5050, right?  It causes a large slowdown
> with some linux guests, as posted here:
>         http://lists.gnu.org/archive/html/qemu-devel/2008-09/msg01201.html
> As mentioned the guest I tested was:
>         sidux-2008-03-ourea-pre1-kde-lite-i386-200809142136.iso
> It uses a Linux 2.6.26 kernel, a FreeBSD guest is not affected by the
> slowness, could this be related to dynticks?
>
>  Oh and on one box transfers over the emulated usb nic even sometimes
> stall, I have to ping the guest's ip for them to continue, and the ping
> then loses packets too.

Can you please clarify your setup. As I understood, from the thread you
mentioned above, you're running Linux 2.6.26 guest on QEMU under FreeBSD 
host,correct ?
And what you're saying is that USB is slow in that configuration.

Assuming that I got that part correctly let me first say that this means 
that usb-linux is not activated and you're using built-in QEMU USB 
devices. What this in turn means is that your USB operation is 
synchronous (all builtin devices are sync). New async logic with 
multiple transactions, etc only applies to the real USB devices 
virtualized via usb-linux & libusb.

After doing UHCI rewrite I made sure that new code generates exactly the 
same UHCI structures as the old code. I may have missed something but in 
general guests should not be affected, and as I explained new async 
logic does not even kick in with built-in devices.
Is there any chance that there may be something else going on here ?
In other words are you sure that it's UHCI changes that cause the slow down.
If you are sure can you please recompile hw/usb-uhci.c with DEBUG 
enabled. Save the trace it generates and send it to me. I can take a 
look at it and see if there is something obvious. I do not have FreeBSD 
hosts around to try and reproduce this myself.

Thanx
Max