Currently in gic_set_affinity() we calculate a temporary cpumask holding
the intersection of the provided cpumask & the CPUs that are online,
then we call cpumask_first twice on it to find the first such CPU. Since
we don't need the temporary cpumask for anything else & we only care
about the first CPU that's both online & in the provided cpumask, we can
instead use cpumask_first_and to find that CPU & drop the temporary
mask.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17110/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
When parsing mti,reserved-cpu-vectors we generate a mask of all bits
that have been declared reserved, the loop through starting from bit 2
to find one that isn't reserved (ie. is zero).
This patch accomplishes the same task more simply by:
- Inititialising the reserved mask to 0x3 (ie. the 2 software
interrupts). This means we don't need to skip them later as the loop
previously has.
- Replacing the loop checking for zero bits with find_first_zero_bit,
which fits our needs now that the 2 software interrupts are marked
reserved. This requires that the type of reserved is changed to
unsigned long so that it's suitable for use with bitmap functions.
- Replacing the magic number 8 with the hamming weight of the ST0_IM
field - ie. the number of bits that a MIPS CPU has for interrupt
inputs. This is still a compile-time constant 8, but makes it
clearer why it's 8.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17054/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
This patch avoids the need to read the GIC_SH_MASK* registers when
decoding shared interrupts by setting & clearing the interrupt's bit in
the appropriate CPU's pcpu_masks entry when masking or unmasking the
interrupt.
This effectively means that whilst an interrupt is masked we clear its
bit in all pcpu_masks, which causes gic_handle_shared_int() to ignore it
on all CPUs without needing to check GIC_SH_MASK*.
In essence, we add a little overhead to masking or unmasking interrupts
but in return reduce the overhead of the far more common task of
decoding interrupts.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17109/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
gic_basic_init() is now a fairly short function that is only called in
one place. Inline it into gic_of_init() to help readability.
[ralf@linux-mips.org: Resolved conflict.]
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17051/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The __gic_init() function is only called from gic_of_init() now that the
non-DT path has been removed. In order to simplify the code & aid
readability, fold __gic_init() into gic_of_init().
This provides us with the ability to return an error code, which
__gic_init() was previously unable to do. As such the irq_domain_add_*()
error paths are modified to print & return an error rather than panic().
[ralf@linux-mips.org: Resoled reject.]
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17050/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Pull irqchip updates for 4.14 from Marc Zyngier:
- irqchip-specific part of the monster GICv4 series
- new UniPhier AIDET irqchip driver
- new variants of some Freescale MSI widget
- blanket removal of of_node->full_name in printk
- random collection of fixes
For LS1046a and LS1043a v1.1, the MSI controller has 4 MSIRs and 4 GIC
SPI interrupts which can be associated with different Core.
So we can support affinity to improve the performance.
The MSI message data is a byte for Layerscape MSI.
7 6 5 4 3 2 1 0
| - | IBS | SRS |
SRS bit0-1 is to select a MSIR which is associated with a CPU.
IBS bit2-6 of ls1046, bit2-4 of ls1043a v1.1 is to select bit of the
MSIR. With affinity, only bits of MSIR0(srs=0 cpu0) are available.
All other bits of the MSIR1-3(cpu1-3) are reserved. The MSI hwirq
always equals bit index of the MSIR0. When changing affinity, MSI
message data will be appended corresponding SRS then MSI will be
moved to the corresponding core.
But in affinity mode, there is only 8 MSI interrupts for a controller
of LS1043a v1.1. It cannot meet the requirement of the some PCIe
devices such as 4 ports Ethernet card. In contrast, without affinity,
all MSIRs can be used for core 0, the MSI interrupts can up to 32.
So the parameter is added to control affinity mode.
"lsmsi=no-affinity" will disable affinity and increase MSI
interrupt number.
Signed-off-by: Minghuan Lian <Minghuan.Lian@nxp.com>
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
A MSI controller of LS1043a v1.0 only includes one MSIR and
is assigned one GIC interrupt. In order to support affinity,
LS1043a v1.1 MSI is assigned 4 MSIRs and 4 GIC interrupts.
But the MSIR has the different offset and only supports 8 MSIs.
The bits between variable bit_start and bit_end in structure
ls_scfg_msir are used to show 8 MSI interrupts. msir_irqs and
msir_base are added to describe the difference of MSI between
LS1043a v1.1 and other SoCs.
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Minghuan Lian <Minghuan.Lian@nxp.com>
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
LS1046a includes 4 MSIRs, each MSIR is assigned a dedicate GIC
SPI interrupt and provides 32 MSI interrupts. Compared to previous
MSI, LS1046a's IBS(interrupt bit select) shift is changed to 2 and
total MSI interrupt number is changed to 128.
The patch adds structure 'ls_scfg_msir' to describe MSIR setting and
'ibs_shift' to store the different value between the SoCs.
Signed-off-by: Minghuan Lian <Minghuan.Lian@nxp.com>
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Initialization of irq_fwd_mask was done using __raw_writel() which
happens to work for all cases except when using ARM BE8 which requires
writel() (with the proper swapping). Move the initialization of the
irq_fwd_mask till later when we have correctly defined our I/O
accessors.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Make these const as they are only used during a copy operation. Done
using Coccinelle.
@match disable optional_qualifier@
identifier s;
@@
static struct mmp_intc_conf s = {...};
@ref@
position p;
identifier match.s;
@@
s@p
@good1@
position ref.p;
identifier match.s,f,c;
expression e;
@@
(
e = s@p
|
e = s@p.f
|
c(...,s@p.f,...)
|
c(...,s@p,...)
)
@bad depends on !good1@
position ref.p;
identifier match.s;
@@
s@p
@depends on forall !bad disable optional_qualifier@
identifier match.s;
@@
static
+ const
struct mmp_intc_conf s;
Signed-off-by: Bhumika Goyal <bhumirks@gmail.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Make this const as it is only used in a copy operation.
Done using Coccinelle.
Signed-off-by: Bhumika Goyal <bhumirks@gmail.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
As KVM needs to know about the availability of GICv4 to enable
direct injection of interrupts, let's advertise the feature in
the gic_kvm_info structure.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
When creating a VM, it is very convenient to have an irq domain
containing all the doorbell interrupts associated with that VM
(each interrupt representing a VPE).
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
A long time ago, GITS_CTLR[1] used to be called GITC_CTLR.EnableVLPI.
It has been subsequently deprecated and is now an "Implementation
Defined" bit that may ot may not be set for GICv4. Brilliant.
And the current crop of the FastModel requires that bit for VLPIs
to be enabled. Oh well... Let's set it and find out what breaks.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
While the doorbell interrupts are usually driven by the HW itself,
having a way to trigger them independently has proved to be a
really useful debug feature. As it is actually very little code,
let's add it to the VPE irqchip operations.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
After moving a VPE from a redistributor to another, we're still left
with a potential pending doorbell interrupt on the old redistributor.
That interrupt should be moved to the new one to be either cleared
or take, depending on what the hypervisor wishes to do.
So let's move it right after having execited VMOVP. This doesn't
add much cost in the !DirectLPI case (we trade a DISCARD for a MOVI),
and the cost of the DIRECTLPI case should be minimal (two extra MMIO
accesses).
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
When we don't have the DirectLPI feature, we must work around the
architecture shortcomings to be able to perform the required
maintenance (interrupt masking, clearing and injection).
For this, we create a fake device whose sole purpose is to
provide a way to issue commands as if we were dealing with LPIs
coming from that device (while they actually originate from
the ITS). This fake device doesn't have LPIs allocated to it,
but instead uses the VPE LPIs.
Of course, this could be a real bottleneck, and a naive
implementation would require 6 commands to issue an invalidation.
Instead, let's allocate at least one event per physical CPU
(rounded up to the next power of 2), and opportunistically
map the VPE doorbell to an event. This doorbell will be mapped
until we roll over and need to reallocate this slot.
This ensures that most of the time, we only need 2 commands
to issue an INV, INT or CLEAR, making the performance a lot
better, given that we always issue a CLEAR on entry, and
an INV on each side of a trapped WFI.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
The normal course of action when allocating the ITS' view of a
device is to allocate the corresponding LPIs. But we're about
to introduce devices that borrow their interrupts from
some other entities.
So let's make the allocation optional.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
When masking/unmasking a doorbell interrupt, it is necessary
to issue an invalidation to the corresponding redistributor.
We use the DirectLPI feature by writting directly to the corresponding
redistributor.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
When we're about to run a vcpu, it is crucial that the redistributor
associated with the physical CPU is being told about the new residency.
This is abstracted by hijacking the irq_set_affinity method for the
doorbell interrupt associated with the VPE. It is expected that the
hypervisor will call this method before scheduling the VPE.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
When a guest issues a INVALL command targetting a collection, it must
be translated into a VINVALL for the VPE that has this collection.
This patch implements a hook that offers this functionallity to the
hypervisor.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
When a VPE is scheduled to run, the corresponding redistributor must
be told so, by setting VPROPBASER to the VM's property table, and
VPENDBASER to the vcpu's pending table.
When scheduled out, we preserve the IDAI and PendingLast bits. The
latter is specially important, as it tells the hypervisor that
there are pending interrupts for this vcpu.
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
On activation, a VPE is mapped using the VMAPP command, followed
by a VINVALL for a good measure. On deactivation, the VPE is
simply unmapped.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
When creating a VM, the low level GICv4 code is responsible for:
- allocating each VPE a unique VPEID
- allocating a doorbell interrupt for each VPE
- allocating the pending tables for each VPE
- allocating the property table for the VM
This of course has to be reversed when the VM is brought down.
All of this is wired into the irq domain alloc/free methods.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
