Skip to contentSkip to conversation

Rationalize max_grant_frames and max_maptrack_frames handling

PATCH Superseded Paul Durrant 3 participants 9 messages Nov 27, 2019 – Nov 28, 2019 Arm/actions semi architecture

v2 of 7 Superseded Latest is v7 →

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 Rationalize max_grant_frames and max_maptrack_frames handling

From: George Dunlap <[email protected]>

Xen used to have single, system-wide limits for the number of grant frames and maptrack frames a guest was allowed to create. Increasing or decreasing this single limit on the Xen command-line would change the limit for all guests on the system.

Later, per-domain limits for these values was created. The system-wide limits became strict limits: domains could not be created with higher limits, but could be created with lower limits. However, that change also introduced a range of different "default" values into various places in the toolstack:

- The python libxc bindings hard-coded these values to 32 and 1024, respectively
- The libxl default values are 32 and 1024 respectively.
- xl will use the libxl default for maptrack, but does its own default calculation for grant frames: either 32 or 64, based on the max possible mfn.

These defaults interact poorly with the hypervisor command-line limit:

- The hypervisor command-line limit cannot be used to raise the limit for all guests anymore, as the default in the toolstack will effectively override this.
- If you use the hypervisor command-line limit to *reduce* the limit, then the "default" values generated by the toolstack are too high, and all guest creations will fail.

In other words, the toolstack defaults require any change to be effected by having the admin explicitly specify a new value in every guest.

In order to address this, have grant_table_init treat negative values for max_grant_frames and max_maptrack_frames as instructions to use the system-wide default, and have all the above toolstacks default to passing -1 unless a different value is explicitly configured.

This restores the old behavior in that changing the hypervisor command-line option can change the behavior for all guests, while retaining the ability to set per-guest values. It also removes the bug that reducing the system-wide max will cause all domains without explicit limits to fail.

NOTE: The Ocaml bindings require the caller to always specify a value, and
the code to start a xenstored stubdomain hard-codes these to 4 and 128 respectively; this behavour will not be modified.

---
Cc: Ian Jackson <[email protected]>
Cc: Wei Liu <[email protected]>
Cc: Andrew Cooper <[email protected]>
Cc: George Dunlap <[email protected]>
Cc: Jan Beulich <[email protected]>
Cc: Julien Grall <[email protected]>
Cc: Konrad Rzeszutek Wilk <[email protected]>
Cc: Stefano Stabellini <[email protected]>
Cc: Anthony PERARD <[email protected]>
Cc: "Marek Marczykowski-Górecki" <[email protected]>
Cc: Volodymyr Babchuk <[email protected]>
Cc: "Roger Pau Monné" <[email protected]>

v2:
- re-worked George's original commit massage a little
- fixed the text in xl.conf.5.pod
- use -1 as the sentinel value for 'default' and < 0 for checking it

---
 docs/man/xl.conf.5.pod            |  6 ++++--
 tools/libxl/libxl.h               |  4 ++--
 tools/libxl/libxl_types.idl       |  4 ++--
 tools/python/xen/lowlevel/xc/xc.c |  4 ++--
 tools/xl/xl.c                     |  8 --------
 tools/xl/xl_parse.c               |  3 ++-
 xen/arch/arm/setup.c              |  2 +-
 xen/arch/x86/setup.c              |  4 ++--
 xen/common/grant_table.c          | 10 ++++++++--
 xen/include/public/domctl.h       | 10 ++++++----
 xen/include/xen/grant_table.h     |  8 ++++----
 11 files changed, 33 insertions(+), 30 deletions(-)
docs/man/xl.conf.5.pod +4 -2
index 962144e38e..207ab3e77a 100644
@@ -81,13 +81,15 @@ Default: C</var/lock/xl>
 
 Sets the default value for the C<max_grant_frames> domain config value.
 
-Default: C<32> on hosts up to 16TB of memory, C<64> on hosts larger than 16TB
+Default: value of Xen command line B<gnttab_max_frames> parameter (or its
+default value if unspecified).
 
 =item B<max_maptrack_frames=NUMBER>
 
 Sets the default value for the C<max_maptrack_frames> domain config value.
 
-Default: C<1024>
+Default: value of Xen command line B<gnttab_max_maptrack_frames>
+parameter (or its default value if unspecified).
 
 =item B<vif.default.script="PATH">
 
tools/libxl/libxl.h +2 -2
index 49b56fa1a3..a2a5d321c5 100644
@@ -364,8 +364,8 @@
  */
 #define LIBXL_HAVE_BUILDINFO_GRANT_LIMITS 1
 
-#define LIBXL_MAX_GRANT_FRAMES_DEFAULT 32
-#define LIBXL_MAX_MAPTRACK_FRAMES_DEFAULT 1024
+#define LIBXL_MAX_GRANT_FRAMES_DEFAULT -1
+#define LIBXL_MAX_MAPTRACK_FRAMES_DEFAULT -1
 
 /*
  * LIBXL_HAVE_BUILDINFO_* indicates that libxl_domain_build_info has
tools/libxl/libxl_types.idl +2 -2
index 0546d7865a..63e29bb2fb 100644
@@ -511,8 +511,8 @@ libxl_domain_build_info = Struct("domain_build_info",[
 
     ("vnuma_nodes", Array(libxl_vnode_info, "num_vnuma_nodes")),
 
-    ("max_grant_frames",    uint32, {'init_val': 'LIBXL_MAX_GRANT_FRAMES_DEFAULT'}),
-    ("max_maptrack_frames", uint32, {'init_val': 'LIBXL_MAX_MAPTRACK_FRAMES_DEFAULT'}),
+    ("max_grant_frames",    integer, {'init_val': 'LIBXL_MAX_GRANT_FRAMES_DEFAULT'}),
+    ("max_maptrack_frames", integer, {'init_val': 'LIBXL_MAX_MAPTRACK_FRAMES_DEFAULT'}),
     
     ("device_model_version", libxl_device_model_version),
     ("device_model_stubdomain", libxl_defbool),
tools/python/xen/lowlevel/xc/xc.c +2 -2
index 44d3606141..a751e85910 100644
@@ -127,8 +127,8 @@ static PyObject *pyxc_domain_create(XcObject *self,
         },
         .max_vcpus = 1,
         .max_evtchn_port = -1, /* No limit. */
-        .max_grant_frames = 32,
-        .max_maptrack_frames = 1024,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
 
     static char *kwd_list[] = { "domid", "ssidref", "handle", "flags",
tools/xl/xl.c +0 -8
index ddd29b3f1b..08f31cc90e 100644
@@ -96,7 +96,6 @@ static void parse_global_config(const char *configfile,
     XLU_Config *config;
     int e;
     const char *buf;
-    libxl_physinfo physinfo;
 
     config = xlu_cfg_init(stderr, configfile);
     if (!config) {
@@ -199,13 +198,6 @@ static void parse_global_config(const char *configfile,
 
     if (!xlu_cfg_get_long (config, "max_grant_frames", &l, 0))
         max_grant_frames = l;
-    else {
-        libxl_physinfo_init(&physinfo);
-        max_grant_frames = (libxl_get_physinfo(ctx, &physinfo) != 0 ||
-                            !(physinfo.max_possible_mfn >> 32))
-                           ? 32 : 64;
-        libxl_physinfo_dispose(&physinfo);
-    }
     if (!xlu_cfg_get_long (config, "max_maptrack_frames", &l, 0))
         max_maptrack_frames = l;
 
tools/xl/xl_parse.c +2 -1
index 112f8ee026..7554048f0b 100644
@@ -1415,9 +1415,10 @@ void parse_config_data(const char *config_source,
         b_info->max_grant_frames = l;
     else
         b_info->max_grant_frames = max_grant_frames;
+
     if (!xlu_cfg_get_long (config, "max_maptrack_frames", &l, 0))
         b_info->max_maptrack_frames = l;
-    else if (max_maptrack_frames != -1)
+    else
         b_info->max_maptrack_frames = max_maptrack_frames;
 
     libxl_defbool_set(&b_info->claim_mode, claim_mode);
xen/arch/arm/setup.c +1 -1
index 51d32106b7..3c899cd4a0 100644
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long boot_phys_offset,
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 
xen/arch/x86/setup.c +2 -2
index 00ee87bde5..7d27f36053 100644
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long mbi_p)
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ? XEN_DOMCTL_CDF_s3_integrity : 0,
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
 
     /* Critical region without IDT or TSS.  Any fault is deadly! */
xen/common/grant_table.c +8 -2
index b34d520f6d..6972cef1de 100644
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int max_grant_frames,
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {
     struct grant_table *gt;
     int ret = -ENOMEM;
 
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||
          max_grant_frames > opt_max_grant_frames ||
          max_maptrack_frames > opt_max_maptrack_frames )
xen/include/public/domctl.h +6 -4
index 9f2cfd602c..e313da499f 100644
@@ -82,13 +82,15 @@ struct xen_domctl_createdomain {
     uint32_t iommu_opts;
 
     /*
-     * Various domain limits, which impact the quantity of resources (global
-     * mapping space, xenheap, etc) a guest may consume.
+     * Various domain limits, which impact the quantity of resources
+     * (global mapping space, xenheap, etc) a guest may consume.  For
+     * max_grant_frames and max_maptrack_frames, < 0 means "use the
+     * default maximum value in the hypervisor".
      */
     uint32_t max_vcpus;
     uint32_t max_evtchn_port;
-    uint32_t max_grant_frames;
-    uint32_t max_maptrack_frames;
+    int32_t max_grant_frames;
+    int32_t max_maptrack_frames;
 
     struct xen_arch_domainconfig arch;
 };
xen/include/xen/grant_table.h +4 -4
index 6f9345d9ef..34886bb6f8 100644
@@ -36,8 +36,8 @@ extern unsigned int opt_max_grant_frames;
 extern unsigned int opt_max_maptrack_frames;
 
 /* Create/destroy per-domain grant table context. */
-int grant_table_init(struct domain *d, unsigned int max_grant_frames,
-                     unsigned int max_maptrack_frames);
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames);
 void grant_table_destroy(
     struct domain *d);
 void grant_table_init_vcpu(struct vcpu *v);
@@ -68,8 +68,8 @@ int gnttab_get_status_frame(struct domain *d, unsigned long idx,
 #define opt_max_maptrack_frames 0
 
 static inline int grant_table_init(struct domain *d,
-                                   unsigned int max_grant_frames,
-                                   unsigned int max_maptrack_frames)
+                                   int max_grant_frames,
+                                   int max_maptrack_frames)
 {
     return 0;
 }
Signature · 7 lines
2.20.1


_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
#1 lore raw
Quoted 14 lines · Paul Durrant
xen/arch/arm/setup.c +1 -1
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long boot_phys_offset,
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long mbi_p) · line 702

Quoted 8 lines
xen/arch/x86/setup.c +2 -2
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long mbi_p)
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ? XEN_DOMCTL_CDF_s3_integrity : 0,
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };

With these there's no need anymore for opt_max_maptrack_frames to be non-static. Sadly Arm still wants opt_max_grant_frames accessible in gnttab_dom0_frames().

Quoted 11 lines
xen/common/grant_table.c +2 -2
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int max_grant_frames,
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {

xen/common/grant_table.c active_alloc_failed: · line 1852

Quoted 10 lines
xen/common/grant_table.c +6 -0
     struct grant_table *gt;
     int ret = -ENOMEM;
 
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||

I take it we don't expect people to specify 2^^31 or more frames for either option. It looks like almost everything here would cope, except for this very comparison. Nevertheless I wonder whether you wouldn't better confine both values to [0, INT_MAX] now, including when adjusted at runtime.

Jan

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel

#2 lore raw
Quoted 43 lines

-----Original Message-----
From: Jan Beulich <[email protected]>
Sent: 27 November 2019 15:56
To: Durrant, Paul <[email protected]>; George Dunlap <[email protected]>
Cc: [email protected]; Andrew Cooper <[email protected]>; Anthony PERARD <[email protected]>; Roger Pau Monné <[email protected]>; Volodymyr Babchuk <[email protected]>; George Dunlap <[email protected]>; Ian Jackson <[email protected]>; Marek Marczykowski-Górecki <[email protected]>; Stefano Stabellini <[email protected]>; Konrad Rzeszutek Wilk <[email protected]>; Julien Grall <[email protected]>; Wei Liu <[email protected]>
Subject: Re: [PATCH v2] Rationalize max_grant_frames and max_maptrack_frames handling

xen/arch/arm/setup.c void __init start_xen(unsigned long

Quoted 3 lines · Paul Durrant
xen/arch/arm/setup.c +0 -0
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long

boot_phys_offset,

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long

Quoted 11 lines
xen/arch/arm/setup.c +1 -1
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 
xen/arch/x86/setup.c +0 -0
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long

mbi_p)

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long · line 698

Quoted 2 lines
xen/arch/x86/setup.c +0 -0
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ? XEN_DOMCTL_CDF_s3_integrity

: 0,

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long mbi_p) · line 702

Quoted 6 lines
xen/arch/x86/setup.c +2 -2
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
Quoted 4 lines

With these there's no need anymore for opt_max_maptrack_frames to be non-static. Sadly Arm still wants opt_max_grant_frames accessible in gnttab_dom0_frames().

Yes, I was about to make them static until I saw what the ARM code did.

Quoted 14 lines
Quoted 14 lines
xen/common/grant_table.c +2 -2
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int max_grant_frames,
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {
     struct grant_table *gt;
     int ret = -ENOMEM;
 
Quoted 13 lines

xen/common/grant_table.c active_alloc_failed: · line 1852

Quoted 7 lines
xen/common/grant_table.c +6 -0
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||

I take it we don't expect people to specify 2^^31 or more frames for either option. It looks like almost everything here would cope, except for this very comparison. Nevertheless I wonder whether you wouldn't better confine both values to [0, INT_MAX] now, including when adjusted at runtime.

I can certainly remove the 'U' from the definition of INITIAL_NR_GRANT_FRAMES, but do you want me to make opt_max_grant_frames and opt_max_maptrack_frames into signed ints and add signed parser code too? I also don't understand the 'adjusted at runtime' part.

Paul

Quoted 2 lines

Jan

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel

#3 lore raw
Quoted 30 lines · Paul
Quoted 30 lines

From: Jan Beulich <[email protected]>
Sent: 27 November 2019 15:56

xen/arch/arm/setup.c void __init start_xen(unsigned long

Quoted 3 lines · Paul Durrant
xen/arch/arm/setup.c +0 -0
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long

boot_phys_offset,

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long

Quoted 11 lines
xen/arch/arm/setup.c +1 -1
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 
xen/arch/x86/setup.c +0 -0
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long

mbi_p)

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long · line 698

Quoted 2 lines
xen/arch/x86/setup.c +0 -0
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ? XEN_DOMCTL_CDF_s3_integrity

: 0,

Quoted 6 lines
xen/arch/x86/setup.c +2 -2
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
Quoted 5 lines
Quoted 3 lines

With these there's no need anymore for opt_max_maptrack_frames to be non-static. Sadly Arm still wants opt_max_grant_frames accessible in gnttab_dom0_frames().

Yes, I was about to make them static until I saw what the ARM code did.

But the one that Arm doesn't need should become static now.

Quoted 28 lines
Quoted 14 lines
Quoted 14 lines
xen/common/grant_table.c +2 -2
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int max_grant_frames,
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {
     struct grant_table *gt;
     int ret = -ENOMEM;
 
Quoted 13 lines

xen/common/grant_table.c active_alloc_failed: · line 1852

Quoted 7 lines
xen/common/grant_table.c +6 -0
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||

I take it we don't expect people to specify 2^^31 or more frames for either option. It looks like almost everything here would cope, except for this very comparison. Nevertheless I wonder whether you wouldn't better confine both values to [0, INT_MAX] now, including when adjusted at runtime.

Quoted 2 lines

I can certainly remove the 'U' from the definition of INITIAL_NR_GRANT_FRAMES,

Oh, I didn't pay attention that is has a U on it - in this case the comparison above is fine.

Quoted 3 lines

but do you want me to make opt_max_grant_frames and opt_max_maptrack_frames into signed ints and add signed parser code too?

Definitely not. They should remain unsigned quantities, but their values may need sanity checking now.

Quoted 1 line

I also don't understand the 'adjusted at runtime' part.

Well, for a command line drive value you could adjust an out of bounds value in some __init function. But for runtime modifiable settings you won't get away this easily.

Jan

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel

#4 lore raw
Quoted 77 lines · Jan Beulich
Quoted 30 lines · Paul
Quoted 30 lines

From: Jan Beulich <[email protected]>
Sent: 27 November 2019 15:56

xen/arch/arm/setup.c void __init start_xen(unsigned long

Quoted 3 lines · Paul Durrant
xen/arch/arm/setup.c +0 -0
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long

boot_phys_offset,

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long

Quoted 11 lines
xen/arch/arm/setup.c +1 -1
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 
xen/arch/x86/setup.c +0 -0
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long

mbi_p)

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long · line 698

Quoted 2 lines
xen/arch/x86/setup.c +0 -0
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ? XEN_DOMCTL_CDF_s3_integrity

: 0,

Quoted 6 lines
xen/arch/x86/setup.c +2 -2
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
Quoted 5 lines
Quoted 3 lines

With these there's no need anymore for opt_max_maptrack_frames to be non-static. Sadly Arm still wants opt_max_grant_frames accessible in gnttab_dom0_frames().

Yes, I was about to make them static until I saw what the ARM code did.

But the one that Arm doesn't need should become static now.

Quoted 28 lines
Quoted 14 lines
Quoted 14 lines
xen/common/grant_table.c +2 -2
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int max_grant_frames,
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {
     struct grant_table *gt;
     int ret = -ENOMEM;
 
Quoted 13 lines

xen/common/grant_table.c active_alloc_failed: · line 1852

Quoted 7 lines
xen/common/grant_table.c +6 -0
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||

I take it we don't expect people to specify 2^^31 or more frames for either option. It looks like almost everything here would cope, except for this very comparison. Nevertheless I wonder whether you wouldn't better confine both values to [0, INT_MAX] now, including when adjusted at runtime.

Quoted 2 lines

I can certainly remove the 'U' from the definition of INITIAL_NR_GRANT_FRAMES,

Oh, I didn't pay attention that is has a U on it - in this case the comparison above is fine.

Quoted 3 lines

but do you want me to make opt_max_grant_frames and opt_max_maptrack_frames into signed ints and add signed parser code too?

Quoted 8 lines

Definitely not. They should remain unsigned quantities, but their values may need sanity checking now.

Quoted 1 line

I also don't understand the 'adjusted at runtime' part.

Well, for a command line drive value you could adjust an out of bounds value in some __init function. But for runtime modifiable settings you won't get away this easily.

TBH I'd be tempted to define XENSOMETHING_MAX_DEFAULT as (unsigned long)(-1) or something, and explicitly compare to that. That leaves open the possibility of having more sentinel values if we decided we wanted them.

-George

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel

#5 lore raw
Quoted 100 lines

-----Original Message-----
From: George Dunlap <[email protected]>
Sent: 27 November 2019 16:34
To: Jan Beulich <[email protected]>; Durrant, Paul <[email protected]>
Cc: AndrewCooper <[email protected]>; Anthony PERARD <[email protected]>; Roger Pau Monné <[email protected]>; Volodymyr Babchuk <[email protected]>; George Dunlap <[email protected]>; Ian Jackson <[email protected]>; Marek Marczykowski-Górecki <[email protected]>; Stefano Stabellini <[email protected]>; [email protected]; Konrad Rzeszutek Wilk <[email protected]>; Julien Grall <[email protected]>; Wei Liu <[email protected]>
Subject: Re: [PATCH v2] Rationalize max_grant_frames and max_maptrack_frames handling

Quoted 17 lines · Jan Beulich
Quoted 16 lines · Paul
Quoted 16 lines

From: Jan Beulich <[email protected]>
Sent: 27 November 2019 15:56

xen/arch/arm/setup.c void __init start_xen(unsigned long

Quoted 3 lines · Paul Durrant
xen/arch/arm/setup.c +0 -0
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long

boot_phys_offset,

Quoted 8 lines
xen/arch/arm/setup.c +1 -1
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long

Quoted 3 lines
xen/arch/x86/setup.c +0 -0
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long

mbi_p)

Quoted 2 lines
xen/arch/x86/setup.c +0 -0
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ?

XEN_DOMCTL_CDF_s3_integrity

Quoted 16 lines
Quoted 8 lines
Quoted 8 lines

: 0,

Quoted 6 lines
xen/arch/x86/setup.c +2 -2
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
Quoted 5 lines
Quoted 3 lines

With these there's no need anymore for opt_max_maptrack_frames to be non-static. Sadly Arm still wants opt_max_grant_frames accessible in gnttab_dom0_frames().

Yes, I was about to make them static until I saw what the ARM code did.

But the one that Arm doesn't need should become static now.

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

xen/common/grant_table.c active_alloc_failed: · line 1840

Quoted 7 lines
xen/common/grant_table.c +0 -1
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int

max_grant_frames,

Quoted 36 lines
Quoted 21 lines
Quoted 7 lines
Quoted 7 lines
xen/common/grant_table.c +2 -1
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {
     struct grant_table *gt;
     int ret = -ENOMEM;
 
Quoted 13 lines

xen/common/grant_table.c active_alloc_failed: · line 1852

Quoted 7 lines
xen/common/grant_table.c +6 -0
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||

I take it we don't expect people to specify 2^^31 or more frames for either option. It looks like almost everything here would cope, except for this very comparison. Nevertheless I wonder whether you wouldn't better confine both values to [0, INT_MAX] now, including when adjusted at runtime.

Quoted 2 lines

I can certainly remove the 'U' from the definition of INITIAL_NR_GRANT_FRAMES,

Oh, I didn't pay attention that is has a U on it - in this case the comparison above is fine.

Quoted 3 lines

but do you want me to make opt_max_grant_frames and opt_max_maptrack_frames into signed ints and add signed parser code too?

Definitely not. They should remain unsigned quantities, but their values may need sanity checking now.

Quoted 1 line

I also don't understand the 'adjusted at runtime' part.

Quoted 8 lines
Quoted 3 lines

Well, for a command line drive value you could adjust an out of bounds value in some __init function. But for runtime modifiable settings you won't get away this easily.

TBH I'd be tempted to define XENSOMETHING_MAX_DEFAULT as (unsigned long)(-1) or something, and explicitly compare to that. That leaves open the possibility of having more sentinel values if we decided we wanted them.

I'm extremely confused now. What do you want me to compare and where?

I assume we're talking about the opt_XXX values. Am I supposed to stop >INT_MAX being assigned to them? Or should I define local unsigned values for max_maptrack/grant_frames and simply initialize them to the passed-in arg (if >= 0) or the opt_XXX value otherwise.

Paul

Quoted 2 lines

-George

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel

#6 lore raw
Quoted 104 lines · Paul
Quoted 104 lines

-----Original Message-----
From: George Dunlap <[email protected]>
Sent: 27 November 2019 16:34
To: Jan Beulich <[email protected]>; Durrant, Paul <[email protected]>
Cc: AndrewCooper <[email protected]>; Anthony PERARD <[email protected]>; Roger Pau Monné <[email protected]>; Volodymyr Babchuk <[email protected]>; George Dunlap <[email protected]>; Ian Jackson <[email protected]>; Marek Marczykowski-Górecki <[email protected]>; Stefano Stabellini <[email protected]>; [email protected]; Konrad Rzeszutek Wilk <[email protected]>; Julien Grall <[email protected]>; Wei Liu <[email protected]>
Subject: Re: [PATCH v2] Rationalize max_grant_frames and max_maptrack_frames handling

Quoted 17 lines · Jan Beulich
Quoted 16 lines · Paul
Quoted 16 lines

From: Jan Beulich <[email protected]>
Sent: 27 November 2019 15:56

xen/arch/arm/setup.c void __init start_xen(unsigned long

Quoted 3 lines · Paul Durrant
xen/arch/arm/setup.c +0 -0
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long

boot_phys_offset,

Quoted 8 lines
xen/arch/arm/setup.c +1 -1
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long

Quoted 3 lines
xen/arch/x86/setup.c +0 -0
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long

mbi_p)

Quoted 2 lines
xen/arch/x86/setup.c +0 -0
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ?

XEN_DOMCTL_CDF_s3_integrity

Quoted 16 lines
Quoted 8 lines
Quoted 8 lines

: 0,

Quoted 6 lines
xen/arch/x86/setup.c +2 -2
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
Quoted 5 lines
Quoted 3 lines

With these there's no need anymore for opt_max_maptrack_frames to be non-static. Sadly Arm still wants opt_max_grant_frames accessible in gnttab_dom0_frames().

Yes, I was about to make them static until I saw what the ARM code did.

But the one that Arm doesn't need should become static now.

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

xen/common/grant_table.c active_alloc_failed: · line 1840

Quoted 7 lines
xen/common/grant_table.c +0 -1
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int

max_grant_frames,

Quoted 39 lines
Quoted 21 lines
Quoted 7 lines
Quoted 7 lines
xen/common/grant_table.c +2 -1
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {
     struct grant_table *gt;
     int ret = -ENOMEM;
 
Quoted 13 lines

xen/common/grant_table.c active_alloc_failed: · line 1852

Quoted 7 lines
xen/common/grant_table.c +6 -0
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||

I take it we don't expect people to specify 2^^31 or more frames for either option. It looks like almost everything here would cope, except for this very comparison. Nevertheless I wonder whether you wouldn't better confine both values to [0, INT_MAX] now, including when adjusted at runtime.

Quoted 2 lines

I can certainly remove the 'U' from the definition of INITIAL_NR_GRANT_FRAMES,

Oh, I didn't pay attention that is has a U on it - in this case the comparison above is fine.

Quoted 3 lines

but do you want me to make opt_max_grant_frames and opt_max_maptrack_frames into signed ints and add signed parser code too?

Definitely not. They should remain unsigned quantities, but their values may need sanity checking now.

Quoted 1 line

I also don't understand the 'adjusted at runtime' part.

Well, for a command line drive value you could adjust an out of bounds value in some __init function. But for runtime modifiable settings you won't get away this easily.

Quoted 8 lines
Quoted 4 lines

TBH I'd be tempted to define XENSOMETHING_MAX_DEFAULT as (unsigned long)(-1) or something, and explicitly compare to that. That leaves open the possibility of having more sentinel values if we decided we wanted them.

I'm extremely confused now. What do you want me to compare and where?

I assume we're talking about the opt_XXX values. Am I supposed to stop >INT_MAX being assigned to them? Or should I define local unsigned values for max_maptrack/grant_frames and simply initialize them to the passed-in arg (if >= 0) or the opt_XXX value otherwise.

In this version of the patch, you change the domctl arguments from uint32_t to int32_t. I would leave them uint32_t, and if ( max_grant_frames == XENSOMETHING_MAX_DEFAULT ) max_grant_frames = opt_&c.

Then the only invalid value we have to worry about is checking for XENSOMETHING_MAX_DEFAULT.

This is a suggestion, and I wouldn't argue strongly if someone thought it was a bad idea, but it seems like the most straightforward option to me.

-George

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel

#7 lore raw
Quoted 138 lines

-----Original Message-----
From: George Dunlap <[email protected]>
Sent: 27 November 2019 16:52
To: Durrant, Paul <[email protected]>; Jan Beulich <[email protected]>
Cc: AndrewCooper <[email protected]>; Anthony PERARD <[email protected]>; Roger Pau Monné <[email protected]>; Volodymyr Babchuk <[email protected]>; George Dunlap <[email protected]>; Ian Jackson <[email protected]>; Marek Marczykowski-Górecki <[email protected]>; Stefano Stabellini <[email protected]>; [email protected]; Konrad Rzeszutek Wilk <[email protected]>; Julien Grall <[email protected]>; Wei Liu <[email protected]>
Subject: Re: [PATCH v2] Rationalize max_grant_frames and max_maptrack_frames handling

Quoted 4 lines · Paul
Quoted 4 lines

-----Original Message-----
From: George Dunlap <[email protected]>
Sent: 27 November 2019 16:34
To: Jan Beulich <[email protected]>; Durrant, Paul

<[email protected]>

Quoted 44 lines
Quoted 44 lines

Cc: AndrewCooper <[email protected]>; Anthony PERARD <[email protected]>; Roger Pau Monné <[email protected]>; Volodymyr Babchuk <[email protected]>; George Dunlap <[email protected]>; Ian Jackson <[email protected]>; Marek Marczykowski-Górecki <[email protected]>; Stefano Stabellini <[email protected]>; [email protected]; Konrad Rzeszutek Wilk <[email protected]>; Julien Grall <[email protected]>; Wei Liu <[email protected]>
Subject: Re: [PATCH v2] Rationalize max_grant_frames and max_maptrack_frames handling

Quoted 17 lines · Jan Beulich
Quoted 16 lines · Paul
Quoted 16 lines

From: Jan Beulich <[email protected]>
Sent: 27 November 2019 15:56

xen/arch/arm/setup.c void __init start_xen(unsigned long

Quoted 3 lines · Paul Durrant
xen/arch/arm/setup.c +0 -0
@@ -789,7 +789,7 @@ void __init start_xen(unsigned long

boot_phys_offset,

Quoted 8 lines
xen/arch/arm/setup.c +1 -1
         .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap,
         .max_evtchn_port = -1,
         .max_grant_frames = gnttab_dom0_frames(),
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_maptrack_frames = -1,
     };
     int rc;
 
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

xen/arch/x86/setup.c void __init noreturn __start_xen(unsigned long

Quoted 3 lines
xen/arch/x86/setup.c +0 -0
@@ -697,8 +697,8 @@ void __init noreturn __start_xen(unsigned long

mbi_p)

Quoted 2 lines
xen/arch/x86/setup.c +0 -0
     struct xen_domctl_createdomain dom0_cfg = {
         .flags = IS_ENABLED(CONFIG_TBOOT) ?

XEN_DOMCTL_CDF_s3_integrity

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

: 0,

Quoted 6 lines
xen/arch/x86/setup.c +2 -2
         .max_evtchn_port = -1,
-        .max_grant_frames = opt_max_grant_frames,
-        .max_maptrack_frames = opt_max_maptrack_frames,
+        .max_grant_frames = -1,
+        .max_maptrack_frames = -1,
     };
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 3 lines

With these there's no need anymore for opt_max_maptrack_frames to be non-static. Sadly Arm still wants opt_max_grant_frames accessible in gnttab_dom0_frames().

Yes, I was about to make them static until I saw what the ARM code

did.

Quoted 56 lines
Quoted 47 lines
Quoted 10 lines

But the one that Arm doesn't need should become static now.

Quoted 7 lines
Quoted 7 lines

xen/common/grant_table.c active_alloc_failed: · line 1840

Quoted 7 lines
xen/common/grant_table.c +0 -1
@@ -1837,12 +1837,18 @@ active_alloc_failed:
     return -ENOMEM;
 }
 
-int grant_table_init(struct domain *d, unsigned int

max_grant_frames,

Quoted 36 lines
Quoted 21 lines
Quoted 7 lines
Quoted 7 lines
xen/common/grant_table.c +2 -1
-                     unsigned int max_maptrack_frames)
+int grant_table_init(struct domain *d, int max_grant_frames,
+                     int max_maptrack_frames)
 {
     struct grant_table *gt;
     int ret = -ENOMEM;
 
Quoted 13 lines

xen/common/grant_table.c active_alloc_failed: · line 1852

Quoted 7 lines
xen/common/grant_table.c +6 -0
+    /* Default to maximum value if no value was specified */
+    if ( max_grant_frames < 0 )
+        max_grant_frames = opt_max_grant_frames;
+    if ( max_maptrack_frames < 0 )
+        max_maptrack_frames = opt_max_maptrack_frames;
+
     if ( max_grant_frames < INITIAL_NR_GRANT_FRAMES ||

I take it we don't expect people to specify 2^^31 or more frames for either option. It looks like almost everything here would cope, except for this very comparison. Nevertheless I wonder whether you wouldn't better confine both values to [0, INT_MAX] now, including when adjusted at runtime.

Quoted 2 lines

I can certainly remove the 'U' from the definition of INITIAL_NR_GRANT_FRAMES,

Oh, I didn't pay attention that is has a U on it - in this case the comparison above is fine.

Quoted 3 lines

but do you want me to make opt_max_grant_frames and opt_max_maptrack_frames into signed ints and add signed parser code too?

Definitely not. They should remain unsigned quantities, but their values may need sanity checking now.

Quoted 1 line

I also don't understand the 'adjusted at runtime' part.

Quoted 8 lines
Quoted 3 lines

Well, for a command line drive value you could adjust an out of bounds value in some __init function. But for runtime modifiable settings you won't get away this easily.

TBH I'd be tempted to define XENSOMETHING_MAX_DEFAULT as (unsigned long)(-1) or something, and explicitly compare to that. That leaves open the possibility of having more sentinel values if we decided we wanted them.

Quoted 4 lines

I'm extremely confused now. What do you want me to compare and where?

I assume we're talking about the opt_XXX values. Am I supposed to stop INT_MAX being assigned to them? Or should I define local unsigned values

for max_maptrack/grant_frames and simply initialize them to the passed-in arg (if >= 0) or the opt_XXX value otherwise.

In this version of the patch, you change the domctl arguments from uint32_t to int32_t. I would leave them uint32_t, and if ( max_grant_frames == XENSOMETHING_MAX_DEFAULT ) max_grant_frames = opt_&c.

Quoted 6 lines

Then the only invalid value we have to worry about is checking for XENSOMETHING_MAX_DEFAULT.

This is a suggestion, and I wouldn't argue strongly if someone thought it was a bad idea, but it seems like the most straightforward option to me.

AFAICT the definition of that invalid value is going to be needed by both the grant table code and the user-space toolstack code so I guess the logical place for the definition would be a tools-only section of the public grant table header? TBH I prefer the idea of any negative value being default though. As long as the xl/libxl parts don't allow a *specified* value > INT_MAX then that should be fine, although for the full story a custom parser for the command line values should also be added to ensure the same semantics there.

Paul

Quoted 2 lines

-George

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel

#8 lore raw
Quoted 30 lines · Paul
Quoted 23 lines

From: George Dunlap <[email protected]>
Sent: 27 November 2019 16:52

Quoted 8 lines · Paul
Quoted 7 lines

From: George Dunlap <[email protected]>
Sent: 27 November 2019 16:34

TBH I'd be tempted to define XENSOMETHING_MAX_DEFAULT as (unsigned long)(-1) or something, and explicitly compare to that. That leaves open the possibility of having more sentinel values if we decided we wanted them.

Quoted 4 lines

I'm extremely confused now. What do you want me to compare and where?

I assume we're talking about the opt_XXX values. Am I supposed to stop INT_MAX being assigned to them? Or should I define local unsigned values

for max_maptrack/grant_frames and simply initialize them to the passed-in arg (if >= 0) or the opt_XXX value otherwise.

In this version of the patch, you change the domctl arguments from uint32_t to int32_t. I would leave them uint32_t, and if ( max_grant_frames == XENSOMETHING_MAX_DEFAULT ) max_grant_frames = opt_&c.

Quoted 6 lines

Then the only invalid value we have to worry about is checking for XENSOMETHING_MAX_DEFAULT.

This is a suggestion, and I wouldn't argue strongly if someone thought it was a bad idea, but it seems like the most straightforward option to me.

Quoted 5 lines

AFAICT the definition of that invalid value is going to be needed by both the grant table code and the user-space toolstack code so I guess the logical place for the definition would be a tools-only section of the public grant table header? TBH I prefer the idea of any negative value being default though.

FWIW I agree, as I can't really see what other purposes we might need sentinel values for down the road.

Quoted 4 lines

As long as the xl/libxl parts don't allow a *specified* value > INT_MAX then that should be fine, although for the full story a custom parser for the command line values should also be added to ensure the same semantics there.

Right. While going a little far, I can't right now see easy alternatives to a custom parser.

Jan

_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel