1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
|
From 0c1e0d59ceb3c4c8909d7a7b6e62d19c7ad81bef Mon Sep 17 00:00:00 2001
From: Radim Krcmar <rkrcmar@redhat.com>
Date: Mon, 10 Mar 2014 15:14:27 +0100
Subject: Workaround for a win8.1-32 S4 resume bug
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Radim Krcmar <rkrcmar@redhat.com>
Message-id: <1394464467-23560-1-git-send-email-rkrcmar@redhat.com>
Patchwork-id: 58069
O-Subject: [RHEL7.0 seabios PATCH] Workaround for a win8.1-32 S4 resume bug
Bugzilla: 1050775
RH-Acked-by: Paolo Bonzini <pbonzini@redhat.com>
RH-Acked-by: Gerd Hoffmann <kraxel@redhat.com>
RH-Acked-by: Laszlo Ersek <lersek@redhat.com>
bug: https://bugzilla.redhat.com/show_bug.cgi?id=1050775
brew: http://brewweb.devel.redhat.com/brew/taskinfo?taskID=7176174
This patch has no upstream equivalent.
When a 32 bit version of windows 8.1 resumes from suspend, it writes 1
into 0x72 in the early boot because it didn't expect a NULL pointer.
0x72 is lower offset byte of 0x1c interrupt entry, so we jump into a
middle of other function if this interrupt is triggered.
Because 0x1c is only triggered from our handle_08, we detect if our
default value (function that does only iret) has its lower offset byte
overwritten and skip it in that case.
(Windows never sets own callback there, so we always detect this bug
correctly, as seabios doesn't use it either
Other sources shouldn't incorrectly overwrite it or use seabios code,
but it is quite ok even if the guest did this on purpose.)
The reason Windows uses NULL pointer is still unknown, but this bug is
blocking WHQL certification, so we have to work around it in 7.0.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
---
(It is either a Windows bug that is going to be solved on their side,
or we will find a better solution after consulting what went wrong.
Happens on RHEL6 too.)
src/clock.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
(cherry picked from commit aa1c26538deecfd820b7da9b3be09ebc20b7fef9)
---
src/clock.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/src/clock.c b/src/clock.c
index e44e112..298a722 100644
--- a/src/clock.c
+++ b/src/clock.c
@@ -309,7 +309,13 @@ handle_08(void)
struct bregs br;
memset(&br, 0, sizeof(br));
br.flags = F_IF;
- call16_int(0x1c, &br);
+ struct segoff_s isr1c = GET_IVT(0x1c);
+ // hardcoded address of entry_iret_official with lower segment byte
+ // overwritten by 1
+ if (isr1c.seg == ((SEG_BIOS & ~0xff) | 0x1) && isr1c.offset == 0xff53)
+ dprintf(1, "Worked around win8.1-32 S4 resume bug\n");
+ else
+ call16_int(0x1c, &br);
pic_eoi1();
}
--
1.8.3.1
|
