From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Jan Beulich <JBeulich@suse.com>
Cc: "Keir (Xen.org)" <keir@xen.org>,
"xiantao.zhang@intel.com" <xiantao.zhang@intel.com>,
xen-devel <xen-devel@lists.xen.org>
Subject: Re: Ping: [PATCH] x86/IO-APIC: fix guest RTE write corner cases
Date: Wed, 15 May 2013 11:57:14 +0100 [thread overview]
Message-ID: <51936A0A.4000009@citrix.com> (raw)
In-Reply-To: <51926BBC02000078000D6148@nat28.tlf.novell.com>
On 14/05/13 15:52, Jan Beulich wrote:
>>>> On 08.05.13 at 14:51, "Jan Beulich" <JBeulich@suse.com> wrote:
>> This fixes two regressions from c/s 20143:a7de5bd776ca ("x86: Make the
>> hypercall PHYSDEVOP_alloc_irq_vector hypercall dummy"):
>>
>> For one, IRQs that had their vector set up by Xen internally without a
>> handler ever having got set (e.g. via "com<n>=..." without a matching
>> consumer option like "console=com<n>") would wrongly call
>> add_pin_to_irq() here, triggering the BUG_ON() in that function.
>>
>> Second, when assign_irq_vector() fails this addition to irq_2_pin[]
>> needs to be undone.
>>
>> In the context of this I'm also surprised that the irq_2_pin[]
>> manipulations here occur without any lock, i.e. rely on Dom0 to do
>> some sort of serialization.
>>
>> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> No-one having any opinion? I'm hesitant to commit changes like
> this without anyone having said any word...
>
> Jan
Sorry - this was on my list to look at but I have been very busy recently.
It tentatively looks good to me.
~Andrew
>
>> --- a/xen/arch/x86/io_apic.c
>> +++ b/xen/arch/x86/io_apic.c
>> @@ -133,6 +133,37 @@ static void add_pin_to_irq(unsigned int
>> share_vector_maps(irq_2_pin[irq].apic, apic);
>> }
>>
>> +static void remove_pin_from_irq(unsigned int irq, int apic, int pin)
>> +{
>> + struct irq_pin_list *entry, *prev;
>> +
>> + for (entry = &irq_2_pin[irq]; ; entry = &irq_2_pin[entry->next]) {
>> + if ((entry->apic == apic) && (entry->pin == pin))
>> + break;
>> + BUG_ON(!entry->next);
>> + }
>> +
>> + entry->pin = entry->apic = -1;
>> +
>> + if (entry != &irq_2_pin[irq]) {
>> + /* Removed entry is not at head of list. */
>> + prev = &irq_2_pin[irq];
>> + while (&irq_2_pin[prev->next] != entry)
>> + prev = &irq_2_pin[prev->next];
>> + prev->next = entry->next;
>> + } else if (entry->next) {
>> + /* Removed entry is at head of multi-item list. */
>> + prev = entry;
>> + entry = &irq_2_pin[entry->next];
>> + *prev = *entry;
>> + entry->pin = entry->apic = -1;
>> + } else
>> + return;
>> +
>> + entry->next = irq_2_pin_free_entry;
>> + irq_2_pin_free_entry = entry - irq_2_pin;
>> +}
>> +
>> /*
>> * Reroute an IRQ to a different pin.
>> */
>> @@ -2280,7 +2311,7 @@ int ioapic_guest_read(unsigned long phys
>>
>> int ioapic_guest_write(unsigned long physbase, unsigned int reg, u32 val)
>> {
>> - int apic, pin, irq, ret, vector, pirq;
>> + int apic, pin, irq, ret, pirq;
>> struct IO_APIC_route_entry rte = { 0 };
>> unsigned long flags;
>> struct irq_desc *desc;
>> @@ -2348,13 +2379,25 @@ int ioapic_guest_write(unsigned long phy
>> return 0;
>> }
>>
>> - if ( desc->arch.vector <= 0 || desc->arch.vector > LAST_DYNAMIC_VECTOR ) {
>> - add_pin_to_irq(irq, apic, pin);
>> - vector = assign_irq_vector(irq, NULL);
>> - if ( vector < 0 )
>> - return vector;
>> + if ( desc->arch.vector <= 0 || desc->arch.vector > LAST_DYNAMIC_VECTOR )
>> + {
>> + int vector = desc->arch.vector;
>> +
>> + if ( vector < FIRST_HIPRIORITY_VECTOR )
>> + add_pin_to_irq(irq, apic, pin);
>> + else
>> + desc->arch.vector = IRQ_VECTOR_UNASSIGNED;
>> + ret = assign_irq_vector(irq, NULL);
>> + if ( ret < 0 )
>> + {
>> + if ( vector < FIRST_HIPRIORITY_VECTOR )
>> + remove_pin_from_irq(irq, apic, pin);
>> + else
>> + desc->arch.vector = vector;
>> + return ret;
>> + }
>>
>> - printk(XENLOG_INFO "allocated vector %02x for irq %d\n", vector, irq);
>> + printk(XENLOG_INFO "allocated vector %02x for irq %d\n", ret, irq);
>> }
>> spin_lock(&dom0->event_lock);
>> ret = map_domain_pirq(dom0, pirq, irq,
>
>
next prev parent reply other threads:[~2013-05-15 10:57 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-05-08 12:51 [PATCH] x86/IO-APIC: fix guest RTE write corner cases Jan Beulich
2013-05-14 14:52 ` Ping: " Jan Beulich
2013-05-14 15:17 ` Keir Fraser
2013-05-15 10:57 ` Andrew Cooper [this message]
2013-05-16 12:47 ` Zhang, Xiantao
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=51936A0A.4000009@citrix.com \
--to=andrew.cooper3@citrix.com \
--cc=JBeulich@suse.com \
--cc=keir@xen.org \
--cc=xen-devel@lists.xen.org \
--cc=xiantao.zhang@intel.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.