On 01/03/2020 11:35, Mark Cave-Ayland wrote:
> On 29/02/2020 23:02, BALATON Zoltan wrote:
>
>> Some machines operate in "non 100% native mode" where interrupts are
>> fixed at legacy IDE interrupts and some guests expect this behaviour
>> without checking based on knowledge about hardware. Even Linux has
>> arch specific workarounds for this that are activated on such boards
>> so this needs to be emulated as well.
>>
>> Signed-off-by: BALATON Zoltan <bala...@eik.bme.hu>
>> ---
>> hw/ide/via.c | 60 +++++++++++++++++++++++++++++++++++------
>> hw/mips/mips_fulong2e.c | 2 +-
>> include/hw/ide.h | 3 ++-
>> 3 files changed, 55 insertions(+), 10 deletions(-)
>>
>> diff --git a/hw/ide/via.c b/hw/ide/via.c
>> index 096de8dba0..17418c5822 100644
>> --- a/hw/ide/via.c
>> +++ b/hw/ide/via.c
>> @@ -1,9 +1,10 @@
>> /*
>> - * QEMU IDE Emulation: PCI VIA82C686B support.
>> + * QEMU VIA southbridge IDE emulation (VT82C686B, VT8231)
>> *
>> * Copyright (c) 2003 Fabrice Bellard
>> * Copyright (c) 2006 Openedhand Ltd.
>> * Copyright (c) 2010 Huacai Chen <zltjiang...@gmail.com>
>> + * Copyright (c) 2019-2020 BALATON Zoltan
>> *
>> * Permission is hereby granted, free of charge, to any person obtaining a
>> copy
>> * of this software and associated documentation files (the "Software"), to
>> deal
>> @@ -25,6 +26,8 @@
>> */
>>
>> #include "qemu/osdep.h"
>> +#include "qemu/range.h"
>> +#include "hw/qdev-properties.h"
>> #include "hw/pci/pci.h"
>> #include "migration/vmstate.h"
>> #include "qemu/module.h"
>> @@ -111,14 +114,43 @@ static void via_ide_set_irq(void *opaque, int n, int
>> level)
>> } else {
>> d->config[0x70 + n * 8] &= ~0x80;
>> }
>> -
>> level = (d->config[0x70] & 0x80) || (d->config[0x78] & 0x80);
>> - n = pci_get_byte(d->config + PCI_INTERRUPT_LINE);
>> - if (n) {
>> - qemu_set_irq(isa_get_irq(NULL, n), level);
>> +
>> + /*
>> + * Some machines operate in "non 100% native mode" where
>> PCI_INTERRUPT_LINE
>> + * is not used but IDE always uses ISA IRQ 14 and 15 even in native
>> mode.
>> + * Some guest drivers expect this, often without checking.
>> + */
>> + if (!(pci_get_byte(d->config + PCI_CLASS_PROG) & (n ? 4 : 1)) ||
>> + PCI_IDE(d)->flags & BIT(PCI_IDE_LEGACY_IRQ)) {
>> + qemu_set_irq(isa_get_irq(NULL, (n ? 15 : 14)), level);
>> + } else {
>> + n = pci_get_byte(d->config + PCI_INTERRUPT_LINE);
>> + if (n) {
>> + qemu_set_irq(isa_get_irq(NULL, n), level);
>> + }
>> }
>> }
The other part I'm not sure about is that I can't see how via_ide_set_irq() can
ever
raise a native PCI IRQ - comparing with my experience on cmd646, should there
not be
a pci_set_irq(d, level) at the end?
ATB,
Mark.
