Re: [PATCH V3 21/23] xen/arm: Add mapcache invalidation handling

From: Oleksandr <olekstysh@gmail.com>
To: Stefano Stabellini <sstabellini@kernel.org>,
	Julien Grall <julien@xen.org>
Cc: xen-devel@lists.xenproject.org,
	Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>,
	Volodymyr Babchuk <Volodymyr_Babchuk@epam.com>,
	Julien Grall <julien.grall@arm.com>
Subject: Re: [PATCH V3 21/23] xen/arm: Add mapcache invalidation handling
Date: Fri, 11 Dec 2020 13:21:15 +0200	[thread overview]
Message-ID: <7c8a9ad9-2b18-7028-17bc-20ee5a341323@gmail.com> (raw)
In-Reply-To: <alpine.DEB.2.21.2012101443060.20986@sstabellini-ThinkPad-T480s>


On 11.12.20 03:28, Stefano Stabellini wrote:

Hi Julien, Stefano

> On Thu, 10 Dec 2020, Julien Grall wrote:
>> On 10/12/2020 02:30, Stefano Stabellini wrote:
>>> On Mon, 30 Nov 2020, Oleksandr Tyshchenko wrote:
>>>> From: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
>>>>
>>>> We need to send mapcache invalidation request to qemu/demu everytime
>>>> the page gets removed from a guest.
>>>>
>>>> At the moment, the Arm code doesn't explicitely remove the existing
>>>> mapping before inserting the new mapping. Instead, this is done
>>>> implicitely by __p2m_set_entry().
>>>>
>>>> So we need to recognize a case when old entry is a RAM page *and*
>>>> the new MFN is different in order to set the corresponding flag.
>>>> The most suitable place to do this is p2m_free_entry(), there
>>>> we can find the correct leaf type. The invalidation request
>>>> will be sent in do_trap_hypercall() later on.
>>> Why is it sent in do_trap_hypercall() ?
>> I believe this is following the approach used by x86. There are actually some
>> discussion about it (see [1]).
>>
>> Leaving aside the toolstack case for now, AFAIK, the only way a guest can
>> modify its p2m is via an hypercall. Do you have an example otherwise?
> OK this is a very important assumption. We should write it down for sure.
> I think it is true today on ARM.
>
>
>> When sending the invalidation request, the vCPU will be blocked until all the
>> IOREQ server have acknowledged the invalidation. So the hypercall seems to be
>> the best position to do it.
>>
>> Alternatively, we could use check_for_vcpu_work() to check if the mapcache
>> needs to be invalidated. The inconvenience is we would execute a few more
>> instructions in each entry/exit path.
> Yeah it would be more natural to call it from check_for_vcpu_work(). If
> we put it between #ifdef CONFIG_IOREQ_SERVER it wouldn't be bad. But I
> am not a fan of increasing the instructions on the exit path either.
>  From this point of view, putting it at the end of do_trap_hypercall is a
> nice trick actually. Let's just make sure it has a good comment on top.
>
>
>>>> Signed-off-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
>>>> CC: Julien Grall <julien.grall@arm.com>
>>>>
>>>> ---
>>>> Please note, this is a split/cleanup/hardening of Julien's PoC:
>>>> "Add support for Guest IO forwarding to a device emulator"
>>>>
>>>> Changes V1 -> V2:
>>>>      - new patch, some changes were derived from (+ new explanation):
>>>>        xen/ioreq: Make x86's invalidate qemu mapcache handling common
>>>>      - put setting of the flag into __p2m_set_entry()
>>>>      - clarify the conditions when the flag should be set
>>>>      - use domain_has_ioreq_server()
>>>>      - update do_trap_hypercall() by adding local variable
>>>>
>>>> Changes V2 -> V3:
>>>>      - update patch description
>>>>      - move check to p2m_free_entry()
>>>>      - add a comment
>>>>      - use "curr" instead of "v" in do_trap_hypercall()
>>>> ---
>>>> ---
>>>>    xen/arch/arm/p2m.c   | 24 ++++++++++++++++--------
>>>>    xen/arch/arm/traps.c | 13 ++++++++++---
>>>>    2 files changed, 26 insertions(+), 11 deletions(-)
>>>>
>>>> diff --git a/xen/arch/arm/p2m.c b/xen/arch/arm/p2m.c
>>>> index 5b8d494..9674f6f 100644
>>>> --- a/xen/arch/arm/p2m.c
>>>> +++ b/xen/arch/arm/p2m.c
>>>> @@ -1,6 +1,7 @@
>>>>    #include <xen/cpu.h>
>>>>    #include <xen/domain_page.h>
>>>>    #include <xen/iocap.h>
>>>> +#include <xen/ioreq.h>
>>>>    #include <xen/lib.h>
>>>>    #include <xen/sched.h>
>>>>    #include <xen/softirq.h>
>>>> @@ -749,17 +750,24 @@ static void p2m_free_entry(struct p2m_domain *p2m,
>>>>        if ( !p2m_is_valid(entry) )
>>>>            return;
>>>>    -    /* Nothing to do but updating the stats if the entry is a
>>>> super-page. */
>>>> -    if ( p2m_is_superpage(entry, level) )
>>>> +    if ( p2m_is_superpage(entry, level) || (level == 3) )
>>>>        {
>>>> -        p2m->stats.mappings[level]--;
>>>> -        return;
>>>> -    }
>>>> +#ifdef CONFIG_IOREQ_SERVER
>>>> +        /*
>>>> +         * If this gets called (non-recursively) then either the entry
>>>> +         * was replaced by an entry with a different base (valid case) or
>>>> +         * the shattering of a superpage was failed (error case).
>>>> +         * So, at worst, the spurious mapcache invalidation might be
>>>> sent.
>>>> +         */
>>>> +        if ( domain_has_ioreq_server(p2m->domain) &&
>>>> +             (p2m->domain == current->domain) &&
>>>> p2m_is_ram(entry.p2m.type) )
>>>> +            p2m->domain->mapcache_invalidate = true;
>>> Why the (p2m->domain == current->domain) check? Shouldn't we set
>>> mapcache_invalidate to true anyway? What happens if p2m->domain !=
>>> current->domain? We wouldn't want the domain to lose the
>>> mapcache_invalidate notification.
>> This is also discussed in [1]. :) The main question is why would a
>> toolstack/device model modify the guest memory after boot?
>>
>> If we assume it does, then the device model would need to pause the domain
>> before modifying the RAM.
>>
>> We also need to make sure that all the IOREQ servers have invalidated
>> the mapcache before the domain run again.
>>
>> This would require quite a bit of work. I am not sure the effort is worth if
>> there are no active users today.
> OK, that explains why we think p2m->domain == current->domain, but why
> do we need to have a check for it right here?
>
> In other words, we don't think it is realistc to get here with
> p2m->domain != current->domain, but let's say that we do somehow. What's
> the best course of action? Probably, set mapcache_invalidate to true and
> possibly print a warning?
>
> Leaving mapcache_invalidate to false doesn't seem to be what we want to
> do?
>
>   
>>>>        BUILD_BUG_ON(NR_hypercalls < ARRAY_SIZE(arm_hypercall_table) );
>>>>    @@ -1459,7 +1460,7 @@ static void do_trap_hypercall(struct cpu_user_regs
>>>> *regs, register_t *nr,
>>>>            return;
>>>>        }
>>>>    -    current->hcall_preempted = false;
>>>> +    curr->hcall_preempted = false;
>>>>          perfc_incra(hypercalls, *nr);
>>>>        call = arm_hypercall_table[*nr].fn;
>>>> @@ -1472,7 +1473,7 @@ static void do_trap_hypercall(struct cpu_user_regs
>>>> *regs, register_t *nr,
>>>>        HYPERCALL_RESULT_REG(regs) = call(HYPERCALL_ARGS(regs));
>>>>      #ifndef NDEBUG
>>>> -    if ( !current->hcall_preempted )
>>>> +    if ( !curr->hcall_preempted )
>>>>        {
>>>>            /* Deliberately corrupt parameter regs used by this hypercall.
>>>> */
>>>>            switch ( arm_hypercall_table[*nr].nr_args ) {
>>>> @@ -1489,8 +1490,14 @@ static void do_trap_hypercall(struct cpu_user_regs
>>>> *regs, register_t *nr,
>>>>    #endif
>>>>          /* Ensure the hypercall trap instruction is re-executed. */
>>>> -    if ( current->hcall_preempted )
>>>> +    if ( curr->hcall_preempted )
>>>>            regs->pc -= 4;  /* re-execute 'hvc #XEN_HYPERCALL_TAG' */
>>>> +
>>>> +#ifdef CONFIG_IOREQ_SERVER
>>>> +    if ( unlikely(curr->domain->mapcache_invalidate) &&
>>>> +         test_and_clear_bool(curr->domain->mapcache_invalidate) )
>>>> +        ioreq_signal_mapcache_invalidate();
>>> Why not just:
>>>
>>> if ( unlikely(test_and_clear_bool(curr->domain->mapcache_invalidate)) )
>>>       ioreq_signal_mapcache_invalidate();
>>>
>> This seems to match the x86 code. My guess is they tried to prevent the cost
>> of the atomic operation if there is no chance mapcache_invalidate is true.
>>
>> I am split whether the first check is worth it. The atomic operation should be
>> uncontended most of the time, so it should be quick. But it will always be
>> slower than just a read because there is always a store involved.
> I am not a fun of optimizations with unclear benefits :-)
>
>
>> On a related topic, Jan pointed out that the invalidation would not work
>> properly if you have multiple vCPU modifying the P2M at the same time.
>>
Thanks to Julien, he explained all bits in detail. Indeed I followed how 
it was done on x86 (place where to send the invalidation request, the 
code to check whether the flag is set, which at first glance, appears 
odd, etc)
and review comments (to latch current into the local variable, and make 
sure that domain sends invalidation request on itself).
Regarding what to do if p2m->domain != current->domain in 
p2m_free_entry(). Probably we could set flag only if guest is paused, 
otherwise just print a warning. Thoughts?


-- 
Regards,

Oleksandr Tyshchenko