From: "Volker Rümelin" <vr_qemu@t-online.de>
To: "Michael S. Tsirkin" <mst@redhat.com>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Philippe Mathieu-Daudé" <philmd@redhat.com>
Cc: qemu-devel@nongnu.org
Subject: [PATCH v2 10/11] pckbd: correctly disable PS/2 communication
Date: Fri, 7 May 2021 20:09:52 +0200 [thread overview]
Message-ID: <20210507180953.8530-10-vr_qemu@t-online.de> (raw)
In-Reply-To: <a898b0d5-7086-9699-ae8b-9524ad319b01@t-online.de>
Currently the PS/2 controller command KBD_CCMD_MOUSE_DISABLE
doesn't disable the PS/2 mouse communication at all, and the
PS/2 controller commands KBD_CCMD_KBD_DISABLE and
KBD_CCMD_KBD_ENABLE disable and enable the keyboard interrupt,
which is very different from what a real PS/2 controller does.
A guest may notice the difference.
Mask out pending data on disabled queues to correctly disable
the PS/2 controller communication.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
---
hw/input/pckbd.c | 29 ++++++++++++++++++++++-------
1 file changed, 22 insertions(+), 7 deletions(-)
diff --git a/hw/input/pckbd.c b/hw/input/pckbd.c
index 3ecc9c74ab..7c476f7a3e 100644
--- a/hw/input/pckbd.c
+++ b/hw/input/pckbd.c
@@ -130,10 +130,12 @@
#define MOUSE_STATUS_ENABLED 0x20
#define MOUSE_STATUS_SCALE21 0x10
-#define KBD_PENDING_KBD 1
-#define KBD_PENDING_AUX 2
+#define KBD_PENDING_KBD_V3 0x01
+#define KBD_PENDING_AUX_V3 0x02
#define KBD_PENDING_CTRL_KBD 0x04
#define KBD_PENDING_CTRL_AUX 0x08
+#define KBD_PENDING_KBD KBD_MODE_DISABLE_KBD /* 0x10 */
+#define KBD_PENDING_AUX KBD_MODE_DISABLE_MOUSE /* 0x20 */
#define KBD_MIGR_TIMER_PENDING 0x1
@@ -163,8 +165,6 @@ typedef struct KBDState {
hwaddr mask;
} KBDState;
-/* XXX: not generating the irqs if KBD_MODE_DISABLE_KBD is set may be
- incorrect, but it avoids having to simulate exact delays */
static void kbd_update_irq_lines(KBDState *s)
{
int irq_kbd_level, irq_mouse_level;
@@ -178,8 +178,7 @@ static void kbd_update_irq_lines(KBDState *s)
irq_mouse_level = 1;
}
} else {
- if ((s->mode & KBD_MODE_KBD_INT) &&
- !(s->mode & KBD_MODE_DISABLE_KBD)) {
+ if (s->mode & KBD_MODE_KBD_INT) {
irq_kbd_level = 1;
}
}
@@ -197,7 +196,7 @@ static void kbd_deassert_irq(KBDState *s)
static uint8_t kbd_pending(KBDState *s)
{
- return s->pending;
+ return s->pending & (~s->mode | ~(KBD_PENDING_KBD | KBD_PENDING_AUX));
}
/* update irq and KBD_STAT_[MOUSE_]OBF */
@@ -357,6 +356,7 @@ static void kbd_write_command(void *opaque, hwaddr addr,
break;
case KBD_CCMD_MOUSE_ENABLE:
s->mode &= ~KBD_MODE_DISABLE_MOUSE;
+ kbd_safe_update_irq(s);
break;
case KBD_CCMD_TEST_MOUSE:
kbd_queue(s, 0x00, 0);
@@ -436,6 +436,9 @@ static void kbd_write_data(void *opaque, hwaddr addr,
switch(s->write_cmd) {
case 0:
ps2_write_keyboard(s->kbd, val);
+ /* sending data to the keyboard reenables PS/2 communication */
+ s->mode &= ~KBD_MODE_DISABLE_KBD;
+ kbd_safe_update_irq(s);
break;
case KBD_CCMD_WRITE_MODE:
s->mode = val;
@@ -462,6 +465,9 @@ static void kbd_write_data(void *opaque, hwaddr addr,
break;
case KBD_CCMD_WRITE_MOUSE:
ps2_write_mouse(s->mouse, val);
+ /* sending data to the mouse reenables PS/2 communication */
+ s->mode &= ~KBD_MODE_DISABLE_MOUSE;
+ kbd_safe_update_irq(s);
break;
default:
break;
@@ -539,7 +545,16 @@ static int kbd_post_load(void *opaque, int version_id)
s->obsrc = s->status & KBD_STAT_OBF ?
(s->status & KBD_STAT_MOUSE_OBF ? KBD_OBSRC_MOUSE : KBD_OBSRC_KBD) :
0;
+ if (s->pending & KBD_PENDING_KBD_V3) {
+ s->pending |= KBD_PENDING_KBD;
+ }
+ if (s->pending & KBD_PENDING_AUX_V3) {
+ s->pending |= KBD_PENDING_AUX;
+ }
}
+ /* clear all unused flags */
+ s->pending &= KBD_PENDING_CTRL_KBD | KBD_PENDING_CTRL_AUX |
+ KBD_PENDING_KBD | KBD_PENDING_AUX;
if (s->migration_flags & KBD_MIGR_TIMER_PENDING) {
kbd_throttle_timeout(s);
}
--
2.26.2
next prev parent reply other threads:[~2021-05-07 18:24 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-07 18:08 [PATCH v2 00/11] PS/2 controller related fixes Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 01/11] ps2: fix mouse stream corruption Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 02/11] ps2: don't raise an interrupt if queue is full Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 03/11] ps2: don't deassert irq twice if queue is empty Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 04/11] pckbd: split out interrupt line changing code Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 05/11] pckbd: don't update OBF flags if KBD_STAT_OBF is set Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 06/11] pckbd: PS/2 keyboard throttle Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 07/11] pckbd: add state variable for interrupt source Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 08/11] pckbd: add controller response queue Volker Rümelin
2021-05-07 18:09 ` [PATCH v2 09/11] pckbd: add function kbd_pending() Volker Rümelin
2021-05-07 21:37 ` Philippe Mathieu-Daudé
2021-05-07 18:09 ` Volker Rümelin [this message]
2021-05-07 18:09 ` [PATCH v2 11/11] pckbd: remove duplicated keyboard and mouse defines Volker Rümelin
2021-05-09 16:03 ` [PATCH v2 00/11] PS/2 controller related fixes Volker Rümelin
2021-05-10 20:41 ` Volker Rümelin
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=20210507180953.8530-10-vr_qemu@t-online.de \
--to=vr_qemu@t-online.de \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=philmd@redhat.com \
--cc=qemu-devel@nongnu.org \
/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).