[PATCH V4 17/24] xen/ioreq: Introduce domain_has_ioreq_server()

Oleksandr Tyshchenko posted 24 patches 1 month, 2 weeks ago

[PATCH V4 17/24] xen/ioreq: Introduce domain_has_ioreq_server()

Posted by Oleksandr Tyshchenko 1 month, 2 weeks ago
From: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>

This patch introduces a helper the main purpose of which is to check
if a domain is using IOREQ server(s).

On Arm the current benefit is to avoid calling vcpu_ioreq_handle_completion()
(which implies iterating over all possible IOREQ servers anyway)
on every return in leave_hypervisor_to_guest() if there is no active
servers for the particular domain.
Also this helper will be used by one of the subsequent patches on Arm.

Signed-off-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
CC: Julien Grall <julien.grall@arm.com>
[On Arm only]
Tested-by: Wei Chen <Wei.Chen@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 RFC -> V1:
   - new patch

Changes V1 -> V2:
   - update patch description
   - guard helper with CONFIG_IOREQ_SERVER
   - remove "hvm" prefix
   - modify helper to just return d->arch.hvm.ioreq_server.nr_servers
   - put suitable ASSERT()s
   - use ASSERT(d->ioreq_server.server[id] ? !s : !!s) in set_ioreq_server()
   - remove d->ioreq_server.nr_servers = 0 from hvm_ioreq_init()

Changes V2 -> V3:
   - update patch description
   - remove ASSERT()s from the helper, add a comment
   - use #ifdef CONFIG_IOREQ_SERVER inside function body
   - use new ASSERT() construction in set_ioreq_server()

Changes V3 -> V4:
   - update patch description
   - drop per-domain variable "nr_servers"
   - reimplement a helper to count the non-NULL entries
   - make the helper out-of-line
---
 xen/arch/arm/traps.c    | 15 +++++++++------
 xen/common/ioreq.c      | 16 ++++++++++++++++
 xen/include/xen/ioreq.h |  2 ++
 3 files changed, 27 insertions(+), 6 deletions(-)

diff --git a/xen/arch/arm/traps.c b/xen/arch/arm/traps.c
index 4a83e1e..35094d8 100644
--- a/xen/arch/arm/traps.c
+++ b/xen/arch/arm/traps.c
@@ -2262,14 +2262,17 @@ static bool check_for_vcpu_work(void)
     struct vcpu *v = current;
 
 #ifdef CONFIG_IOREQ_SERVER
-    bool handled;
+    if ( domain_has_ioreq_server(v->domain) )
+    {
+        bool handled;
 
-    local_irq_enable();
-    handled = vcpu_ioreq_handle_completion(v);
-    local_irq_disable();
+        local_irq_enable();
+        handled = vcpu_ioreq_handle_completion(v);
+        local_irq_disable();
 
-    if ( !handled )
-        return true;
+        if ( !handled )
+            return true;
+    }
 #endif
 
     if ( likely(!v->arch.need_flush_to_ram) )
diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c
index d5f4dd3..59f4990 100644
--- a/xen/common/ioreq.c
+++ b/xen/common/ioreq.c
@@ -80,6 +80,22 @@ static ioreq_t *get_ioreq(struct ioreq_server *s, struct vcpu *v)
     return &p->vcpu_ioreq[v->vcpu_id];
 }
 
+/*
+ * This should only be used when d == current->domain or when they're
+ * distinct and d is paused. Otherwise the result is stale before
+ * the caller can inspect it.
+ */
+bool domain_has_ioreq_server(const struct domain *d)
+{
+    const struct ioreq_server *s;
+    unsigned int id;
+
+    FOR_EACH_IOREQ_SERVER(d, id, s)
+        return true;
+
+    return false;
+}
+
 static struct ioreq_vcpu *get_pending_vcpu(const struct vcpu *v,
                                            struct ioreq_server **srvp)
 {
diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h
index ec7e98d..f0908af 100644
--- a/xen/include/xen/ioreq.h
+++ b/xen/include/xen/ioreq.h
@@ -81,6 +81,8 @@ static inline bool ioreq_needs_completion(const ioreq_t *ioreq)
 #define HANDLE_BUFIOREQ(s) \
     ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF)
 
+bool domain_has_ioreq_server(const struct domain *d);
+
 bool vcpu_ioreq_pending(struct vcpu *v);
 bool vcpu_ioreq_handle_completion(struct vcpu *v);
 bool is_ioreq_server_page(struct domain *d, const struct page_info *page);
-- 
2.7.4


RE: [PATCH V4 17/24] xen/ioreq: Introduce domain_has_ioreq_server()

Posted by Paul Durrant 1 month, 1 week ago
> -----Original Message-----
> From: Xen-devel <xen-devel-bounces@lists.xenproject.org> On Behalf Of Oleksandr Tyshchenko
> Sent: 12 January 2021 21:52
> To: xen-devel@lists.xenproject.org
> Cc: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>; Stefano Stabellini <sstabellini@kernel.org>;
> Julien Grall <julien@xen.org>; Volodymyr Babchuk <Volodymyr_Babchuk@epam.com>; Paul Durrant
> <paul@xen.org>; Julien Grall <julien.grall@arm.com>
> Subject: [PATCH V4 17/24] xen/ioreq: Introduce domain_has_ioreq_server()
> 
> From: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
> 
> This patch introduces a helper the main purpose of which is to check
> if a domain is using IOREQ server(s).
> 
> On Arm the current benefit is to avoid calling vcpu_ioreq_handle_completion()
> (which implies iterating over all possible IOREQ servers anyway)
> on every return in leave_hypervisor_to_guest() if there is no active
> servers for the particular domain.
> Also this helper will be used by one of the subsequent patches on Arm.
> 
> Signed-off-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
> CC: Julien Grall <julien.grall@arm.com>
> [On Arm only]
> Tested-by: Wei Chen <Wei.Chen@arm.com>
> 

Reviewed-by: Paul Durrant <paul@xen.org>


Re: [PATCH V4 17/24] xen/ioreq: Introduce domain_has_ioreq_server()

Posted by Stefano Stabellini 1 month, 1 week ago
On Tue, 12 Jan 2021, Oleksandr Tyshchenko wrote:
> From: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
> 
> This patch introduces a helper the main purpose of which is to check
> if a domain is using IOREQ server(s).
> 
> On Arm the current benefit is to avoid calling vcpu_ioreq_handle_completion()
> (which implies iterating over all possible IOREQ servers anyway)
> on every return in leave_hypervisor_to_guest() if there is no active
> servers for the particular domain.
> Also this helper will be used by one of the subsequent patches on Arm.
> 
> Signed-off-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
> CC: Julien Grall <julien.grall@arm.com>
> [On Arm only]
> Tested-by: Wei Chen <Wei.Chen@arm.com>

Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>


> ---
> Please note, this is a split/cleanup/hardening of Julien's PoC:
> "Add support for Guest IO forwarding to a device emulator"
> 
> Changes RFC -> V1:
>    - new patch
> 
> Changes V1 -> V2:
>    - update patch description
>    - guard helper with CONFIG_IOREQ_SERVER
>    - remove "hvm" prefix
>    - modify helper to just return d->arch.hvm.ioreq_server.nr_servers
>    - put suitable ASSERT()s
>    - use ASSERT(d->ioreq_server.server[id] ? !s : !!s) in set_ioreq_server()
>    - remove d->ioreq_server.nr_servers = 0 from hvm_ioreq_init()
> 
> Changes V2 -> V3:
>    - update patch description
>    - remove ASSERT()s from the helper, add a comment
>    - use #ifdef CONFIG_IOREQ_SERVER inside function body
>    - use new ASSERT() construction in set_ioreq_server()
> 
> Changes V3 -> V4:
>    - update patch description
>    - drop per-domain variable "nr_servers"
>    - reimplement a helper to count the non-NULL entries
>    - make the helper out-of-line
> ---
>  xen/arch/arm/traps.c    | 15 +++++++++------
>  xen/common/ioreq.c      | 16 ++++++++++++++++
>  xen/include/xen/ioreq.h |  2 ++
>  3 files changed, 27 insertions(+), 6 deletions(-)
> 
> diff --git a/xen/arch/arm/traps.c b/xen/arch/arm/traps.c
> index 4a83e1e..35094d8 100644
> --- a/xen/arch/arm/traps.c
> +++ b/xen/arch/arm/traps.c
> @@ -2262,14 +2262,17 @@ static bool check_for_vcpu_work(void)
>      struct vcpu *v = current;
>  
>  #ifdef CONFIG_IOREQ_SERVER
> -    bool handled;
> +    if ( domain_has_ioreq_server(v->domain) )
> +    {
> +        bool handled;
>  
> -    local_irq_enable();
> -    handled = vcpu_ioreq_handle_completion(v);
> -    local_irq_disable();
> +        local_irq_enable();
> +        handled = vcpu_ioreq_handle_completion(v);
> +        local_irq_disable();
>  
> -    if ( !handled )
> -        return true;
> +        if ( !handled )
> +            return true;
> +    }
>  #endif
>  
>      if ( likely(!v->arch.need_flush_to_ram) )
> diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c
> index d5f4dd3..59f4990 100644
> --- a/xen/common/ioreq.c
> +++ b/xen/common/ioreq.c
> @@ -80,6 +80,22 @@ static ioreq_t *get_ioreq(struct ioreq_server *s, struct vcpu *v)
>      return &p->vcpu_ioreq[v->vcpu_id];
>  }
>  
> +/*
> + * This should only be used when d == current->domain or when they're
> + * distinct and d is paused. Otherwise the result is stale before
> + * the caller can inspect it.
> + */
> +bool domain_has_ioreq_server(const struct domain *d)
> +{
> +    const struct ioreq_server *s;
> +    unsigned int id;
> +
> +    FOR_EACH_IOREQ_SERVER(d, id, s)
> +        return true;
> +
> +    return false;
> +}
> +
>  static struct ioreq_vcpu *get_pending_vcpu(const struct vcpu *v,
>                                             struct ioreq_server **srvp)
>  {
> diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h
> index ec7e98d..f0908af 100644
> --- a/xen/include/xen/ioreq.h
> +++ b/xen/include/xen/ioreq.h
> @@ -81,6 +81,8 @@ static inline bool ioreq_needs_completion(const ioreq_t *ioreq)
>  #define HANDLE_BUFIOREQ(s) \
>      ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF)
>  
> +bool domain_has_ioreq_server(const struct domain *d);
> +
>  bool vcpu_ioreq_pending(struct vcpu *v);
>  bool vcpu_ioreq_handle_completion(struct vcpu *v);
>  bool is_ioreq_server_page(struct domain *d, const struct page_info *page);
> -- 
> 2.7.4
>