qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Laurent Vivier <laurent@vivier.eu>
To: qemu-devel@nongnu.org
Cc: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>,
	Laurent Vivier <laurent@vivier.eu>
Subject: [PULL 15/20] mac_via: workaround NetBSD ADB bus enumeration issue
Date: Sun,  8 Oct 2023 08:23:44 +0200	[thread overview]
Message-ID: <20231008062349.2733552-16-laurent@vivier.eu> (raw)
In-Reply-To: <20231008062349.2733552-1-laurent@vivier.eu>

From: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>

NetBSD assumes it can send its first ADB command after sending the ADB_BUSRESET
command in ADB_STATE_NEW without changing the state back to ADB_STATE_IDLE
first as detailed in the ADB protocol.

Add a workaround to detect this condition at the start of ADB enumeration
and send the next command written to SR after a ADB_BUSRESET onto the bus
regardless, even if we don't detect a state transition to ADB_STATE_NEW.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-ID: <20231004083806.757242-16-mark.cave-ayland@ilande.co.uk>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
---
 hw/misc/mac_via.c    | 34 ++++++++++++++++++++++++++++++++++
 hw/misc/trace-events |  1 +
 2 files changed, 35 insertions(+)

diff --git a/hw/misc/mac_via.c b/hw/misc/mac_via.c
index f2ccb12d643b..9f9c2815d0e9 100644
--- a/hw/misc/mac_via.c
+++ b/hw/misc/mac_via.c
@@ -1001,6 +1001,8 @@ static void mos6522_q800_via1_write(void *opaque, hwaddr addr, uint64_t val,
 {
     MOS6522Q800VIA1State *v1s = MOS6522_Q800_VIA1(opaque);
     MOS6522State *ms = MOS6522(v1s);
+    int oldstate, state;
+    int oldsr = ms->sr;
 
     addr = (addr >> 9) & 0xf;
 
@@ -1016,6 +1018,38 @@ static void mos6522_q800_via1_write(void *opaque, hwaddr addr, uint64_t val,
 
         v1s->last_b = ms->b;
         break;
+
+    case VIA_REG_SR:
+        {
+            /*
+             * NetBSD assumes it can send its first ADB command after sending
+             * the ADB_BUSRESET command in ADB_STATE_NEW without changing the
+             * state back to ADB_STATE_IDLE first as detailed in the ADB
+             * protocol.
+             *
+             * Add a workaround to detect this condition at the start of ADB
+             * enumeration and send the next command written to SR after a
+             * ADB_BUSRESET onto the bus regardless, even if we don't detect a
+             * state transition to ADB_STATE_NEW.
+             *
+             * Note that in my tests the NetBSD state machine takes one ADB
+             * operation to recover which means the probe for an ADB device at
+             * address 1 always fails. However since the first device is at
+             * address 2 then this will work fine, without having to come up
+             * with a more complicated and invasive solution.
+             */
+            oldstate = (v1s->last_b & VIA1B_vADB_StateMask) >>
+                       VIA1B_vADB_StateShift;
+            state = (ms->b & VIA1B_vADB_StateMask) >> VIA1B_vADB_StateShift;
+
+            if (oldstate == ADB_STATE_NEW && state == ADB_STATE_NEW &&
+                    (ms->acr & VIA1ACR_vShiftOut) &&
+                    oldsr == 0 /* ADB_BUSRESET */) {
+                trace_via1_adb_netbsd_enum_hack();
+                adb_via_send(v1s, state, ms->sr);
+            }
+        }
+        break;
     }
 }
 
diff --git a/hw/misc/trace-events b/hw/misc/trace-events
index 5a998f5e4e38..24ba7cc4d0e9 100644
--- a/hw/misc/trace-events
+++ b/hw/misc/trace-events
@@ -271,6 +271,7 @@ via1_rtc_cmd_pram_sect_write(int sector, int offset, int addr, int value) "secto
 via1_adb_send(const char *state, uint8_t data, const char *vadbint) "state %s data=0x%02x vADBInt=%s"
 via1_adb_receive(const char *state, uint8_t data, const char *vadbint, int status, int index, int size) "state %s data=0x%02x vADBInt=%s status=0x%x index=%d size=%d"
 via1_adb_poll(uint8_t data, const char *vadbint, int status, int index, int size) "data=0x%02x vADBInt=%s status=0x%x index=%d size=%d"
+via1_adb_netbsd_enum_hack(void) "using NetBSD enum hack"
 via1_auxmode(int mode) "setting auxmode to %d"
 via1_timer_hack_state(int state) "setting timer_hack_state to %d"
 
-- 
2.41.0



  parent reply	other threads:[~2023-10-08  6:27 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-08  6:23 [PULL 00/20] Q800 for 8.2 patches Laurent Vivier
2023-10-08  6:23 ` [PULL 01/20] q800-glue.c: convert to Resettable interface Laurent Vivier
2023-10-08  6:23 ` [PULL 02/20] q800: add djMEMC memory controller Laurent Vivier
2023-10-08  6:23 ` [PULL 03/20] q800: add machine id register Laurent Vivier
2023-10-08  6:23 ` [PULL 04/20] q800: implement additional machine id bits on VIA1 port A Laurent Vivier
2023-10-08  6:23 ` [PULL 05/20] q800: add IOSB subsystem Laurent Vivier
2023-10-08  6:23 ` [PULL 06/20] q800: allow accesses to RAM area even if less memory is available Laurent Vivier
2023-10-08  6:23 ` [PULL 07/20] audio: add Apple Sound Chip (ASC) emulation Laurent Vivier
2023-10-08  6:23 ` [PULL 08/20] asc: generate silence if FIFO empty but engine still running Laurent Vivier
2023-10-08  6:23 ` [PULL 09/20] q800: add Apple Sound Chip (ASC) audio to machine Laurent Vivier
2023-10-08  6:23 ` [PULL 10/20] q800: add easc bool machine class property to switch between ASC and EASC Laurent Vivier
2023-10-08  6:23 ` [PULL 11/20] swim: add trace events for IWM and ISM registers Laurent Vivier
2023-10-08  6:23 ` [PULL 12/20] swim: split into separate IWM and ISM register blocks Laurent Vivier
2023-10-08  6:23 ` [PULL 13/20] swim: update IWM/ISM register block decoding Laurent Vivier
2023-10-08  6:23 ` [PULL 14/20] mac_via: work around underflow in TimeDBRA timing loop in SETUPTIMEK Laurent Vivier
2023-10-08  6:23 ` Laurent Vivier [this message]
2023-10-08  6:23 ` [PULL 16/20] mac_via: implement ADB_STATE_IDLE state if shift register in input mode Laurent Vivier
2023-10-08  6:23 ` [PULL 17/20] mac_via: always clear ADB interrupt when switching to A/UX mode Laurent Vivier
2023-10-08  6:23 ` [PULL 18/20] q800: add ESCC alias at 0xc000 Laurent Vivier
2023-10-08  6:23 ` [PULL 19/20] q800: add alias for MacOS toolbox ROM at 0x40000000 Laurent Vivier
2023-10-08  6:23 ` [PULL 20/20] mac_via: extend timer calibration hack to work with A/UX Laurent Vivier
2023-10-10  1:29 ` [PULL 00/20] Q800 for 8.2 patches Stefan Hajnoczi

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=20231008062349.2733552-16-laurent@vivier.eu \
    --to=laurent@vivier.eu \
    --cc=mark.cave-ayland@ilande.co.uk \
    --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).