From: "Philippe Mathieu-Daudé" <philmd@redhat.com>
To: "Volker Rümelin" <vr_qemu@t-online.de>,
"Gerd Hoffmann" <kraxel@redhat.com>
Cc: Peter Maydell <peter.maydell@linaro.org>, qemu-devel@nongnu.org
Subject: [PATCH-for-6.1? 0/3] ps2: Fix issue #501 and #502
Date: Sat, 7 Aug 2021 16:29:32 +0200 [thread overview]
Message-ID: <df850743-8e86-dc7b-1006-d63b86b7ee36@redhat.com> (raw)
In-Reply-To: <4d1c8467-d976-2c0f-ba54-c767df7b8fe7@t-online.de>
On 8/7/21 2:10 PM, Volker Rümelin wrote:
> Since commit ff6e1624b3 (pckbd: don't update OBF flags if
> KBD_STAT_OBF is set) the OSes Minoca OS and Visopsys no longer
> have a working PS/2 keyboard and mouse. This is caused by a
> PS/2 queue stall due to a lost interrupt in the guest OS. This
> already happened before commit ff6e1624b3, but no one noticed
> because up to that point QEMU sent gratuitous keyboard and mouse
> interrupts and the queue restarted with keyboard input or mouse
> movement.
>
> The lost interrupt is a guest bug. The fact that it's always
> lost is due to an inexact PS/2 keyboard emulation. The way in
> which the two operating systems e.g. set the keyboard LEDs,
> leaves a keyboard ACK reply in the keyboard queue that is
> unexpected for the guests.
>
> This patch series improves the PS/2 keyboard emulation.
>
> There's a workaround for issue #501 and #502 so I don't think
> this is rc3 material. But that decision is up to the maintainers.
Meanwhile, what about reverting ff6e1624b3 for 6.1?
> To verify patch 2/3 I plugged in an additional PS/2 keyboard
> into the host and started Linux with the command line option
> initcall_blacklist=i8042_init. Here is an example of the sequence
> to set the keyboard LEDs.
>
> # #regular sequence to set the keyboard LEDs
> # inb --hex 0x64
> 1c
> # #PS/2 queue is empty
> # outb 0x60 0xed
> # inb --hex 0x64
> 15
> # inb --hex 0x60
> fa
> # inb --hex 0x64
> 14
> # outb 0x60 0x01
> # inb --hex 0x64
> 15
> # inb --hex 0x60
> fa
> # inb --hex 0x64
> 14
>
> # #alternative sequence to set the keyboard LEDs
> # inb --hex 0x64
> 14
> # outb 0x60 0xed
> # outb 0x60 0x01
> # inb --hex 0x64
> 15
> # inb --hex 0x60
> fa
> # inb --hex 0x64
> 14
>
> Volker Rümelin (3):
> ps2: use the whole ps2 buffer but keep queue size
> ps2: use a separate keyboard command reply queue
> ps2: migration support for command reply queue
>
> hw/input/ps2.c | 214 ++++++++++++++++++++++++++++++-------------------
> 1 file changed, 133 insertions(+), 81 deletions(-)
>
next prev parent reply other threads:[~2021-08-07 14:30 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-07 12:10 ps2: Fix issue #501 and #502 Volker Rümelin
2021-08-07 12:12 ` [PATCH 1/3] ps2: use the whole ps2 buffer but keep queue size Volker Rümelin
2021-08-07 12:12 ` [PATCH 2/3] ps2: use a separate keyboard command reply queue Volker Rümelin
2021-08-09 9:52 ` Gerd Hoffmann
2021-08-07 12:12 ` [PATCH 3/3] ps2: migration support for " Volker Rümelin
2021-08-09 10:18 ` Gerd Hoffmann
2021-08-10 5:05 ` Volker Rümelin
2021-08-10 5:40 ` Gerd Hoffmann
2021-08-10 8:38 ` Volker Rümelin
2021-08-07 14:29 ` Philippe Mathieu-Daudé [this message]
2021-08-10 9:07 ` [PATCH-for-6.1? 0/3] ps2: Fix issue #501 and #502 Gerd Hoffmann
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=df850743-8e86-dc7b-1006-d63b86b7ee36@redhat.com \
--to=philmd@redhat.com \
--cc=kraxel@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=vr_qemu@t-online.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).