When reserved-memory regions are present in the host device tree, dom0
is started with multiple memory nodes. Each memory node should have a
unique name, but today they are all called "memory" leading to Linux
printing the following warning at boot:
OF: Duplicate name in base, renamed to "memory#1"
This patch fixes the problem by appending a "@<unit-address>" to the
name, as per the Device Tree specification, where <unit-address> matches
the base of address of the first region.
Fixes: 248faa637d (xen/arm: add reserved-memory regions to the dom0 memory node)
Reported-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@xilinx.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Call make_memory_node for reserved_memory only if we actually have any
reserved_memory regions to handle.
Add a check in make_memory_node to return an error if it has been called
with no memory banks as argument.
Fixes: 248faa637d (xen/arm: add reserved-memory regions to the dom0 memory node)
Signed-off-by: Stefano Stabellini <stefano.stabellini@xilinx.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Specifically
* xen.org to xenproject.org
* http to https
* Replaced pages where page has moved
(including on xen pages as well as external pages)
* Removed some URLs (e.g. downloads for Linux PV drivers)
Tested-by: Lars Kurth <lars.kurth@citrix.com>
Signed-off-by: Lars Kurth <lars.kurth@citrix.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Wei Liu <wl@xen.org>
Release-acked-by: Juergen Gross <jgross@suse.com>
The credit scheduler is the only scheduler with tick_suspend and
tick_resume callbacks. Today those callbacks are invoked without being
guarded by the scheduler lock which is critical when at the same the
cpu those callbacks are active is being moved to or from a cpupool.
Crashes like the following are possible due to that race:
(XEN) ----[ Xen-4.13.0-8.0.12-d x86_64 debug=y Not tainted ]----
(XEN) CPU: 79
(XEN) RIP: e008:[<ffff82d0802467dc>] set_timer+0x39/0x1f7
(XEN) RFLAGS: 0000000000010002 CONTEXT: hypervisor
<snip>
(XEN) Xen call trace:
(XEN) [<ffff82d0802467dc>] set_timer+0x39/0x1f7
(XEN) [<ffff82d08022c1f4>]
sched_credit.c#csched_tick_resume+0x54/0x59
(XEN) [<ffff82d080241dfe>] sched_tick_resume+0x67/0x86
(XEN) [<ffff82d0802eda52>] mwait-idle.c#mwait_idle+0x32b/0x357
(XEN) [<ffff82d08027939e>] domain.c#idle_loop+0xa6/0xc2
(XEN)
(XEN) Pagetable walk from 0000000000000048:
(XEN) L4[0x000] = 00000082cfb9c063 ffffffffffffffff
(XEN) L3[0x000] = 00000082cfb9b063 ffffffffffffffff
(XEN) L2[0x000] = 00000082cfb9a063 ffffffffffffffff
(XEN) L1[0x000] = 0000000000000000 ffffffffffffffff
(XEN)
(XEN) ****************************************
(XEN) Panic on CPU 79:
(XEN) FATAL PAGE FAULT
(XEN) [error_code=0000]
(XEN) Faulting linear address: 0000000000000048
(XEN) ****************************************
The callbacks are used when the cpu is going to or coming from idle in
order to allow higher C-states.
The credit scheduler knows when it is going to schedule an idle
scheduling unit or another one after idle, so it can easily stop or
resume the timer itself removing the need to do so via the callback.
As this timer handling is done in the main scheduling function the
scheduler lock is still held, so the race with cpupool operations can
no longer occur. Note that calling the callbacks from schedule_cpu_rm()
and schedule_cpu_add() is no longer needed, as the transitions to and
from idle in the cpupool with credit active will automatically occur
and do the right thing.
With the last user of the callbacks gone those can be removed.
Suggested-by: George Dunlap <george.dunlap@citrix.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
xmalloc_array() may return NULL if there are memory. Rather than trying
to deference it directly, we should check the return value first.
Coverity-ID: 1381852
Signed-off-by: Julien Grall <julien.grall@arm.com>
Acked-by: Daniel De Graaf <dgdegra@tycho.nsa.gov>
Release-acked-by: Juergen Gross <jgross@suse.com>
flask_assign_{, dt}device() may be used to check whether you can test if
a device is assigned. In this case, the domain will be NULL.
However, flask_iommu_resource_use_perm() will be called and may end up
to deference a NULL pointer. This can be prevented by moving the call
after we check the validity for the domain pointer.
Coverity-ID: 1486741
Fixes: 71e617a6b8 ('use is_iommu_enabled() where appropriate...')
Signed-off-by: Julien Grall <julien.grall@arm.com>
Acked-by: Daniel De Graaf <dgdegra@tycho.nsa.gov>
Reviewed-by: Paul Durrant <paul@xen.org>
Release-acked-by: Juergen Gross <jgross@suse.com>
This is a minor UI change, but users which have elected to enable
XEN_GUEST (which still defaults to no) will definitely need one of these
options, and will typically want both.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Wei Liu <wl@xen.org>
Acked-by: Jan Beulich <jbeulich@suse.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
There are legitimate circumstance where array hardening is not wanted or
needed. Allow it to be turned off.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
The names in retpoline_safe() are copied from should_use_eager_fpu(). The
names in mds_calculations() come partly from Linux's intel-family.h, and
partly from conversations with Intel.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
We don't have a clear way to know how many virtual SPIs we need for the
dom0-less domains. Introduce a new option under xen,domain to specify
the number of SPIs to allocate for a domain.
The property is optional. When absent, we'll use the physical number of
GIC lines for dom0-less domains, or GUEST_VPL011_SPI+1 if vpl011 is
requested, whichever is greater.
Remove the old setting of nr_spis based on the presence of the vpl011.
The implication of this change is that without nr_spis dom0less domains
get the same amount of SPI allocated as dom0, regardless of how many
physical devices they have assigned, and regardless of whether they have
a virtual pl011 (which also needs an emulated SPI). This is done because
the SPIs allocation needs to be done before parsing any passthrough
information, so we have to account for any potential physical SPI
assigned to the domain.
When nr_spis is present, the domain gets exactly nr_spis allocated SPIs.
If the number is too low, it might not be enough for the devices
assigned it to it. If the number is less than GUEST_VPL011_SPI, the
virtual pl011 won't work.
Signed-off-by: Stefano Stabellini <stefanos@xilinx.com>
Reviewed-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Detect "multiboot,device-tree" compatible nodes. Add them to the bootmod
array as BOOTMOD_GUEST_DTB. In kernel_probe, find the right
BOOTMOD_GUEST_DTB and store a pointer to it in dtb_bootmodule.
Signed-off-by: Stefano Stabellini <stefanos@xilinx.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Scan the user provided dtb fragment at boot. For each device node, map
memory to guests, and route interrupts and setup the iommu.
The memory region to remap is specified by the "xen,reg" property.
The iommu is setup by passing the node of the device to assign on the
host device tree. The path is specified in the device tree fragment as
the "xen,path" string property.
The interrupts are remapped based on the information from the
corresponding node on the host device tree. Call
handle_device_interrupts to remap interrupts. Interrupts related device
tree properties are copied from the device tree fragment, same as all
the other properties.
Require both xen,reg and xen,path to be present, unless
xen,force-assign-without-iommu is also set. In that case, tolerate a
missing xen,path, also tolerate iommu setup failure for the passthrough
device.
Also set add the new flag XEN_DOMCTL_CDF_iommu so that dom0less domU
can use the IOMMU if a partial dtb is specified.
Signed-off-by: Stefano Stabellini <stefanos@xilinx.com>
Reviewed-by: Julien Grall <julien.grall@arm.com>
Read the dtb fragment corresponding to a passthrough device from memory
at the location referred to by the "multiboot,device-tree" compatible
node.
Add a new field named dtb_bootmodule to struct kernel_info to keep track
of the dtb fragment location.
Copy the fragment to the guest dtb (only /aliases and /passthrough).
Set kinfo->phandle_gic based on the phandle of the special "/gic"
node in the device tree fragment. "/gic" is a dummy node in the dtb
fragment that represents the gic interrupt controller. Other properties
in the dtb fragment might refer to it (for instance interrupt-parent of
a device node). We reuse the phandle of "/gic" from the dtb fragment as
the phandle of the full GIC node that will be created for the guest
device tree. That way, when we copy properties from the device tree
fragment to the domU device tree the links remain unbroken.
scan_passthrough_prop is introduced here and not used in this patch but
it will be used by later patches.
Some of the code below is taken from tools/libxl/libxl_arm.c. Note that
it is OK to take LGPL 2.1 code and including it into a GPLv2 code base.
The result is GPLv2 code.
Signed-off-by: Stefano Stabellini <stefanos@xilinx.com>
Acked-by: Julien Grall <julien.grall@arm.com>
----
Changes in v6:
- code style
- in-code comment
- commit message improvements
Changes in v5:
- code style
- in-code comment
- remove depth parameter from scan_pfdt_node
- for instead of loop in domain_handle_dtb_bootmodule
- move "gic" check to domain_handle_dtb_bootmodule
- add check_partial_fdt
- use DT_ROOT_NODE_ADDR/SIZE_CELLS_DEFAULT
- add scan_passthrough_prop parameter, set it to false for "/aliases"
Changes in v4:
- use recursion in the implementation
- rename handle_properties to handle_prop_pfdt
- rename scan_pt_node to scan_pfdt_node
- pass kinfo to handle_properties
- use uint32_t instead of u32
- rename r to res
- add "passthrough" and "aliases" check
- add a name == NULL check
- code style
- move DTB fragment scanning earlier, before DomU GIC node creation
- set guest_phandle_gic based on "/gic"
- in-code comment
Changes in v3:
- switch to using device_tree_for_each_node for the copy
Changes in v2:
- add a note about the code coming from libxl in the commit message
- copy /aliases
- code style
Instead of always hard-coding the GIC phandle (GUEST_PHANDLE_GIC), store
it in a variable under kinfo. This way it can be dynamically chosen per
domain. Remove the fdt pointer argument to the make_*_domU_node
functions and oass a struct kernel_info * instead. The fdt pointer can
be accessed from kinfo->fdt. Remove the struct domain *d parameter to
the make_*_domU_node functions because it becomes unused.
Initialize phandle_gic to GUEST_PHANDLE_GIC at the beginning of
prepare_dtb_domU for DomUs. Later patches will change the value of
phandle_gic depending on user provided information.
For Dom0, initialize phandle_gic to dt_interrupt_controller->phandle
(current value) at the beginning of prepare_dtb.
Signed-off-by: Stefano Stabellini <stefanos@xilinx.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Move the interrupt handling code out of handle_device to a new function
so that it can be reused for dom0less VMs (it will be used in later
patches).
Signed-off-by: Stefano Stabellini <stefanos@xilinx.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Some Device Trees may expose both legacy SMMU and generic IOMMU bindings
together. However, the SMMU driver in Xen is only supporting the legacy
SMMU bindings, leading to fatal initialization errors at boot time.
This patch fixes the booting problem by adding a check to
iommu_add_dt_device: if the Xen driver doesn't support the new generic
bindings, and the device is behind an IOMMU, do not return error. The
following iommu_assign_dt_device should succeed.
This check will become superfluous, hence removable, once the Xen SMMU
driver gets support for the generic IOMMU bindings.
Signed-off-by: Stefano Stabellini <stefano.stabellini@xilinx.com>
Reviewed-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Stubdomain do not have it's own config file - its configuration is
derived from target domains. Do not try to manipulate it when attaching
PCI device.
This bug prevented starting HVM with stubdomain and PCI passthrough
device attached.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
When qemu is running in stubdomain, handling "pci-ins" command will fail
if pcifront is not initialized already. Fix this by sending such command
only after confirming that pciback/front is running.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Reviewed-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
HVM domains use IOMMU and device model assistance for communicating with
PCI devices, xen-pcifront/pciback isn't directly needed by HVM domain.
But pciback serve also second function - it reset the device when it is
deassigned from the guest and for this reason pciback needs to be used
with HVM domain too.
When HVM domain has device model in stubdomain, attaching xen-pciback to
the target domain itself may prevent attaching xen-pciback to the
(PV) stubdomain, effectively breaking PCI passthrough.
Fix this by attaching pciback only to one domain: if PV stubdomain is in
use, let it be stubdomain (the commit prevents attaching device to target
HVM in this case); otherwise, attach it to the target domain.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
When libxl__device_pci_add() is called, stubdomain is already running,
even when still constructing the target domain. Previously, do_pci_add()
was called with 'starting' hardcoded to false, but now do_pci_add() shares
'starting' flag in pci_add_state for both target domain and stubdomain.
Fix this by resetting (local) 'starting' to false in pci_add_dm_done()
(previously part of do_pci_add()) when handling stubdomain, regardless
of pas->starting value.
Fixes: 11db56f9a6 (libxl_pci: Use libxl__ao_device with libxl__device_pci_add)
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Reviewed-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
XED commit 1b2fd94425 ("Update MOVSXD to modern behavior") points out
that as of SDM rev 064 MOVSXD is specified to read only 16 bits from
memory (or register) when used without REX.W and with operand size
override. Since the upper 16 bits of the value read won't be used
anyway in this case, make the emulation uniformly follow this more
compatible behavior when not emulating an AMD-like CPU, at the risk
of missing an exception when emulating on/for older hardware (the
boundary at SandyBridge noted in said commit looks questionable - I've
observed the "new" behavior also on Westmere, and a discussion there
lead to Mark finding that even Merom has this behavior already).
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Before the problematic commit, libxl used to ignore error when
destroying (force == true) a passthrough device. If the DM failed to
detach the pci device within the allowed time, the timed out error
raised skip part of pci_remove_*, but also raise the error up to the
caller of libxl__device_pci_destroy_all, libxl__destroy_domid, and
thus the destruction of the domain fails.
When a *pci_destroy* function is called (so we have force=true), error
should mostly be ignored. If the DM didn't confirmed that the device
is removed, we will print a warning and keep going if force=true.
The patch reorder the functions so that pci_remove_timeout() calls
pci_remove_detatched() like it's done when DM calls are successful.
We also clean the QMP states and associated timeouts earlier, as soon
as they are not needed anymore.
Reported-by: Sander Eikelenboom <linux@eikelenboom.it>
Fixes: fae4880c45
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Tested-by: Sander Eikelenboom <linux@eikelenboom.it>
Release-acked-by: Juergen Gross <jgross@suse.com>
Add a scheduling granularity enum ("cpu", "core", "socket") for
specification of the scheduling granularity. Initially it is set to
"cpu", this can be modified by the new boot parameter (x86 only)
"sched-gran".
According to the selected granularity sched_granularity is set after
all cpus are online.
A test is added for all sched resources holding the same number of
cpus. Fall back to core- or cpu-scheduling in that case.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
When entering deep sleep states all domains are paused resulting in
all cpus only running idle vcpus. This enables us to stop scheduling
completely in order to avoid synchronization problems with core
scheduling when individual cpus are offlined.
Disabling the scheduler is done by replacing the softirq handler
with a dummy scheduling routine only enabling tasklets to run.
Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
With core scheduling active it is necessary to move multiple cpus at
the same time to or from a cpupool in order to avoid split scheduling
resources in between.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
With core scheduling active schedule_cpu_[add/rm]() has to cope with
different scheduling granularity: a cpu not in any cpupool is subject
to granularity 1 (cpu scheduling), while a cpu in a cpupool might be
in a scheduling resource with more than one cpu.
Handle that by having arrays of old/new pdata and vdata and loop over
those where appropriate.
Additionally the scheduling resource(s) must either be merged or
split.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
Prepare supporting multiple cpus per scheduling resource by allocating
the cpumask per resource dynamically.
Modify sched_res_mask to have only one bit per scheduling resource set.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
In order to be able to move cpus to cpupools with core scheduling
active it is mandatory to merge multiple cpus into one scheduling
resource or to split a scheduling resource with multiple cpus in it
into multiple scheduling resources. This in turn requires to modify
the cpu <-> scheduling resource relation. In order to be able to free
unused resources protect struct sched_resource via RCU. This ensures
there are no users left when freeing such a resource.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
Instead of letting schedule_cpu_switch() handle moving cpus from and
to cpupools, split it into schedule_cpu_add() and schedule_cpu_rm().
This will allow us to drop allocating/freeing scheduler data for free
cpus as the idle scheduler doesn't need such data.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
On- and offlining cpus with core scheduling is rather complicated as
the cpus are taken on- or offline one by one, but scheduling wants them
rather to be handled per core.
As the future plan is to be able to select scheduling granularity per
cpupool prepare that by storing the granularity in struct
sched_resource (we need it there for free cpus which are not
associated to any cpupool). Free cpus will always use granularity 1.
Store the selected granularity option (cpu, core or socket) in the
cpupool , as we will need it to select the appropriate cpu mask when
populating the cpupool with cpus.
This will make on- and offlining of cpus much easier and avoids
writing code which would needed to be thrown away later.
Move the granularity related variables to cpupool.c as they are now
used form there only.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
When core or socket scheduling are active enabling or disabling smt is
not possible as that would require a major host reconfiguration.
Add a bool sched_disable_smt_switching which will be set for core or
socket scheduling.
Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Dario Faggioli <dfaggioli@suse.com>
Having a pointer to struct cpupool in struct sched_resource instead
of per cpu is enough.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
Having a pointer to struct scheduler in struct sched_resource instead
of per cpu is enough.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
vcpu_wake() and vcpu_sleep() need to be made core scheduling aware:
they might need to switch a single vcpu of an already scheduled unit
between running and not running.
Especially when vcpu_sleep() for a vcpu is being called by a vcpu of
the same scheduling unit special care must be taken in order to avoid
a deadlock: the vcpu to be put asleep must be forced through a
context switch without doing so for the calling vcpu. For this
purpose add a vcpu flag handled in sched_slave() and in
sched_wait_rendezvous_in() allowing a vcpu of the currently running
unit to switch state at a higher priority than a normal schedule
event.
Use the same mechanism when waking up a vcpu of a currently active
unit.
While at it make vcpu_sleep_nosync_locked() static as it is used in
schedule.c only.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
Acked-by: Julien Grall <julien.grall@arm.com>
When scheduling an unit with multiple vcpus there is no guarantee all
vcpus are available (e.g. above maxvcpus or vcpu offline). Fall back to
idle vcpu of the current cpu in that case. This requires to store the
correct schedule_unit pointer in the idle vcpu as long as it used as
fallback vcpu.
In order to modify the runstates of the correct vcpus when switching
schedule units merge sched_unit_runstate_change() into
sched_switch_units() and loop over the affected physical cpus instead
of the unit's vcpus. This in turn requires an access function to the
current variable of other cpus.
Today context_saved() is called in case previous and next vcpus differ
when doing a context switch. With an idle vcpu being capable to be a
substitute for an offline vcpu this is problematic when switching to
an idle scheduling unit. An idle previous vcpu leaves us in doubt which
schedule unit was active previously, so save the previous unit pointer
in the per-schedule resource area. If it is NULL the unit has not
changed and we don't have to set the previous unit to be not running.
When running an idle vcpu in a non-idle scheduling unit use a specific
guest idle loop not performing any non-softirq tasklets and
livepatching in order to avoid populating the cpu caches with memory
used by other domains (as far as possible). Softirqs are considered to
be save.
In order to avoid livepatching when going to guest idle another
variant of reset_stack_and_jump() not calling check_for_livepatch_work
is needed.
Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Julien Grall <julien.grall@arm.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Add a percpu variable holding the index of the cpu in the current
sched_resource structure. This index is used to get the correct vcpu
of a sched_unit on a specific cpu.
For now this index will be zero for all cpus, but with core scheduling
it will be possible to have higher values, too.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
With a scheduling granularity greater than 1 multiple vcpus share the
same struct sched_unit. Support that.
Setting the initial processor must be done carefully: we can't use
sched_set_res() as that relies on for_each_sched_unit_vcpu() which in
turn needs the vcpu already as a member of the domain's vcpu linked
list, which isn't the case.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
cpupool_domain_cpumask() is used by scheduling to select cpus or to
iterate over cpus. In order to support scheduling units spanning
multiple cpus rename cpupool_domain_cpumask() to
cpupool_domain_master_cpumask() and let it return a cpumask with only
one bit set per scheduling resource.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
In several places there is support for multiple vcpus per sched unit
missing. Add that missing support (with the exception of initial
allocation) and missing helpers for that.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Today the vcpu runstate of a new scheduled vcpu is always set to
"running" even if at that time vcpu_runnable() is already returning
false due to a race (e.g. with pausing the vcpu).
With core scheduling this can no longer work as not all vcpus of a
schedule unit have to be "running" when being scheduled. So the vcpu's
new runstate has to be selected at the same time as the runnability of
the related schedule unit is probed.
For this purpose introduce a new helper unit_runnable_state() which
will save the new runstate of all tested vcpus in a new field of the
vcpu struct.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
When switching sched units synchronize all vcpus of the new unit to be
scheduled at the same time.
A variable sched_granularity is added which holds the number of vcpus
per schedule unit.
As tasklets require to schedule the idle unit it is required to set the
tasklet_work_scheduled parameter of do_schedule() to true if any cpu
covered by the current schedule() call has any pending tasklet work.
For joining other vcpus of the schedule unit we need to add a new
softirq SCHED_SLAVE_SOFTIRQ in order to have a way to initiate a
context switch without calling the generic schedule() function
selecting the vcpu to switch to, as we already know which vcpu we
want to run. This has the other advantage not to loose any other
concurrent SCHEDULE_SOFTIRQ events.
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Julien Grall <julien.grall@arm.com>
The "TO BE DOCUMENTED" section of the xl man page still references
tmem. So does the xl.conf man page. Remove the references.
Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Wei Liu <wl@xen.org>
Release-acked-by: Juergen Gross <jgross@suse.com>
Clang will throw an error if a function is unused unless you tell
to ignore it. This can be done using __maybe_unused.
While modifying the declaration, update it to match prototype of similar
functions (see build_assertions). This helps to understand that the sole
purpose of the function is to hold BUILD_BUG_ON().
Signed-off-by: Julien Grall <julien.grall@arm.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Acked-by: Stefano Stabellini <sstabellini@kernel.org>
Clang will throw an error if a function is unused unless you tell
to ignore it. This can be done using __maybe_unused.
While modifying the declaration, update it to match prototype of similar
functions (see build_assertions). This helps to understand that the sole
purpose of the function is to hold BUILD_BUG_ON().
Signed-off-by: Julien Grall <julien.grall@arm.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Acked-by: Stefano Stabellini <sstabellini@kernel.org>
Clang is pickier than GCC for the register size in asm statement. It
expects the register size to match the value size.
The asm statement expects a 32-bit (resp. 64-bit) value on Arm32
(resp. Arm64) whereas the value is a boolean (Clang consider to be
32-bit).
It would be possible to impose 32-bit register for both architecture
but this require the code to use __OP32. However, it does no really
improve the assembly generated. Instead, replace switch the variable to
use register_t.
Signed-off-by: Julien Grall <julien.grall@arm.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Acked-by: Stefano Stabellini <sstabellini@kernel.org>
Clang is pickier than GCC for the register size in asm statement. It
expects the register size to match the value size.
The asm statement expects a 32-bit (resp. 64-bit) value on Arm32
(resp. Arm64) whereas the value is a boolean (Clang consider to be
32-bit).
It would be possible to impose 32-bit register for both architecture
but this require the code to use __OP32. However, it does not really
improve the assembly generated. Instead, replace switch the variable
to use register_t.
Signed-off-by: Julien Grall <julien.grall@arm.com>
Release-acked-by: Juergen Gross <jgross@suse.com>
Acked-by: Stefano Stabellini <sstabellini@kernel.org>
Stefano Stabellini
Lars Kurth
Juergen Gross
Julien Grall
Andrew Cooper
Marek Marczykowski-Górecki
Jan Beulich
Anthony PERARD