On Sat, 2018-05-12 at 11:19 +0800, Qiu Wenbo wrote:
> Yes. The situation can be easily reproduced by the patch below. We
> have noticed it usually happen on some thin clients (without the
> patch). Also, I can reproduce it on my desktop (latest version of
> Arch linux, ryzen 1600x) with debug build
If I remember correctly, this asynchronous disconnection was needed
because in some case, during normal usage, the disconnection was
blocking for seconds (on Windows with usbdk), and if it's not async, it
would block the main loop, and thus the whole GUI (all of this should
have been in the commit
Yes. The situation can be easily reproduced by the patch below. We have
noticed it usually happen on some thin clients (without the patch).
Also, I can reproduce it on my desktop (latest version of Arch linux,
ryzen 1600x) with debug build of qemu, spice-gtk, and usbredir with
remote-viewer --s
On Mon, 2018-05-07 at 17:02 +0800, Qiu Wenbo wrote:
> This reverts commit 9fbf679453d8dbfe797a738cb536136599d7adab.
>
> In some cases, remote-viewer will exit before the async function run
> in
> another thread finish and USB devices redirected to the VM will not
> "pop up" to
> operation system.
This reverts commit 9fbf679453d8dbfe797a738cb536136599d7adab.
In some cases, remote-viewer will exit before the async function run in
another thread finish and USB devices redirected to the VM will not "pop up" to
operation system. For example, a USB disk should be auto mounted
when remote-viewer
This reverts commit 9fbf679453d8dbfe797a738cb536136599d7adab.
In some cases, remote-viewer will exit before the async function run in
another thread finish and USB devices redirected to the VM will not "pop up" to
operation system. For example, a USB disk should be auto mounted
when remote-viewer
