Re-use the code introduced in pseries to save and dump the contents
of the SLB in the case of an SLB involved machine check exception.
This patch also avoids allocating the SLB save array on pseries radix.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190802105709.27696-9-npiggin@gmail.com
Bare metal machine checks run an "early" handler in real mode before
running the main handler which reports the event.
The main handler runs exactly as a normal interrupt handler, after the
"windup" which sets registers back as they were at interrupt entry.
CFAR does not get restored by the windup code, so that will be wrong
when the handler is run.
Restore the CFAR to the saved value before running the late handler.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190802105709.27696-8-npiggin@gmail.com
This label has only one caller, so unwind the branch and move it
inline. The location of the comment is adjusted to match similar
one in system reset.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190802105709.27696-7-npiggin@gmail.com
Now that pseries with fwnmi registered runs the early machine check
handler, there is no good reason to special case the non-fwnmi case
and skip the early handler. Reducing the code and number of paths is
a top priority for asm code, it's better to handle this in C where
possible (and the pseries early handler is a no-op if fwnmi is not
registered).
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190802105709.27696-6-npiggin@gmail.com
The host kernel delivery case for powernv does RFI_TO_USER_OR_KERNEL,
but should just use RFI_TO_KERNEL which makes it clear this is not a
user case.
This is not a bug because RFI_TO_USER_OR_KERNEL deals with kernel
returns just fine.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190802105709.27696-5-npiggin@gmail.com
The machine_check_handle_early hypervisor guest test is skipped if
!HVMODE or MSR[HV]=0, which is wrong for PR or nested hypervisors
that could be running a guest in this state.
Test HSTATE_IN_GUEST up front and use that to branch out to the KVM
handler, then MSR[PR] alone can test for this kernel's userspace.
This matches all other interrupt handling.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190802105709.27696-4-npiggin@gmail.com
Enables running as a secure guest in platforms with an Ultravisor.
Signed-off-by: Ryan Grimm <grimm@linux.vnet.ibm.com>
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-17-bauerman@linux.ibm.com
SWIOTLB checks range of incoming CPU addresses to be bounced and sees if
the device can access it through its DMA window without requiring bouncing.
In such cases it just chooses to skip bouncing. But for cases like secure
guests on powerpc platform all addresses need to be bounced into the shared
pool of memory because the host cannot access it otherwise. Hence the need
to do the bouncing is not related to device's DMA window and use of bounce
buffers is forced by setting swiotlb_force.
Also, connect the shared memory conversion functions into the
ARCH_HAS_MEM_ENCRYPT hooks and call swiotlb_update_mem_attributes() to
convert SWIOTLB's memory pool to shared memory.
Signed-off-by: Anshuman Khandual <khandual@linux.vnet.ibm.com>
[ bauerman: Use ARCH_HAS_MEM_ENCRYPT hooks to share swiotlb memory pool. ]
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-15-bauerman@linux.ibm.com
Secure guest memory is inacessible to devices so regular DMA isn't
possible.
In that case set devices' dma_map_ops to NULL so that the generic
DMA code path will use SWIOTLB to bounce buffers for DMA.
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-14-bauerman@linux.ibm.com
Normally, the HV emulates some instructions like MSGSNDP, MSGCLRP
from a KVM guest. To emulate the instructions, it must first read
the instruction from the guest's memory and decode its parameters.
However for a secure guest (aka SVM), the page containing the
instruction is in secure memory and the HV cannot access directly.
It would need the Ultravisor (UV) to facilitate accessing the
instruction and parameters but the UV currently does not have
the support for such accesses.
Until the UV has such support, disable doorbells in SVMs. This might
incur a performance hit but that is yet to be quantified.
With this patch applied (needed only in SVMs not needed for HV) we
are able to launch SVM guests with multi-core support. Eg:
qemu -smp sockets=2,cores=2,threads=2.
Fix suggested by Benjamin Herrenschmidt. Thanks to input from
Paul Mackerras, Ram Pai and Michael Anderson.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-13-bauerman@linux.ibm.com
User space might want to know it's running in a secure VM. It can't do
a mfmsr because mfmsr is a privileged instruction.
The solution here is to create a cpu attribute:
/sys/devices/system/cpu/svm
which will read 0 or 1 based on the S bit of the current CPU.
Signed-off-by: Ryan Grimm <grimm@linux.vnet.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-12-bauerman@linux.ibm.com
A new kernel deserves a clean slate. Any pages shared with the hypervisor
is unshared before invoking the new kernel. However there are exceptions.
If the new kernel is invoked to dump the current kernel, or if there is a
explicit request to preserve the state of the current kernel, unsharing
of pages is skipped.
NOTE: While testing crashkernel, make sure at least 256M is reserved for
crashkernel. Otherwise SWIOTLB allocation will fail and crash kernel will
fail to boot.
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-11-bauerman@linux.ibm.com
Secure guests need to share the DTL buffers with the hypervisor. To that
end, use a kmem_cache constructor which converts the underlying buddy
allocated SLUB cache pages into shared memory.
Signed-off-by: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-10-bauerman@linux.ibm.com
LPPACA structures need to be shared with the host. Hence they need to be in
shared memory. Instead of allocating individual chunks of memory for a
given structure from memblock, a contiguous chunk of memory is allocated
and then converted into shared memory. Subsequent allocation requests will
come from the contiguous chunk which will be always shared memory for all
structures.
While we are able to use a kmem_cache constructor for the Debug Trace Log,
LPPACAs are allocated very early in the boot process (before SLUB is
available) so we need to use a simpler scheme here.
Introduce helper is_svm_platform() which uses the S bit of the MSR to tell
whether we're running as a secure guest.
Signed-off-by: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-9-bauerman@linux.ibm.com
Helps document what the hard-coded number means.
Also take the opportunity to fix an #endif comment.
Suggested-by: Alexey Kardashevskiy <aik@linux.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-8-bauerman@linux.ibm.com
Protected Execution Facility (PEF) is an architectural change for
POWER 9 that enables Secure Virtual Machines (SVMs). When enabled,
PEF adds a new higher privileged mode, called Ultravisor mode, to
POWER architecture.
The hardware changes include the following:
* There is a new bit in the MSR that determines whether the current
process is running in secure mode, MSR(S) bit 41. MSR(S)=1, process
is in secure mode, MSR(s)=0 process is in normal mode.
* The MSR(S) bit can only be set by the Ultravisor.
* HRFID cannot be used to set the MSR(S) bit. If the hypervisor needs
to return to a SVM it must use an ultracall. It can determine if
the VM it is returning to is secure.
* The privilege of a process is now determined by three MSR bits,
MSR(S, HV, PR). In each of the tables below the modes are listed
from least privilege to highest privilege. The higher privilege
modes can access all the resources of the lower privilege modes.
**Secure Mode MSR Settings**
+---+---+---+---------------+
| S | HV| PR|Privilege |
+===+===+===+===============+
| 1 | 0 | 1 | Problem |
+---+---+---+---------------+
| 1 | 0 | 0 | Privileged(OS)|
+---+---+---+---------------+
| 1 | 1 | 0 | Ultravisor |
+---+---+---+---------------+
| 1 | 1 | 1 | Reserved |
+---+---+---+---------------+
**Normal Mode MSR Settings**
+---+---+---+---------------+
| S | HV| PR|Privilege |
+===+===+===+===============+
| 0 | 0 | 1 | Problem |
+---+---+---+---------------+
| 0 | 0 | 0 | Privileged(OS)|
+---+---+---+---------------+
| 0 | 1 | 0 | Hypervisor |
+---+---+---+---------------+
| 0 | 1 | 1 | Problem (HV) |
+---+---+---+---------------+
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
[ cclaudio: Update the commit message ]
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-7-bauerman@linux.ibm.com
These functions are used when the guest wants to grant the hypervisor
access to certain pages.
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-6-bauerman@linux.ibm.com
Make the Enter-Secure-Mode (ESM) ultravisor call to switch the VM to secure
mode. Pass kernel base address and FDT address so that the Ultravisor is
able to verify the integrity of the VM using information from the ESM blob.
Add "svm=" command line option to turn on switching to secure mode.
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
[ andmike: Generate an RTAS os-term hcall when the ESM ucall fails. ]
Signed-off-by: Michael Anderson <andmike@linux.ibm.com>
[ bauerman: Cleaned up the code a bit. ]
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-5-bauerman@linux.ibm.com
For secure VMs, the signing tool will create a ticket called the "ESM blob"
for the Enter Secure Mode ultravisor call with the signatures of the kernel
and initrd among other things.
This adds support to the wrapper script for adding that blob via the "-e"
option to the zImage.pseries.
It also adds code to the zImage wrapper itself to retrieve and if necessary
relocate the blob, and pass its address to Linux via the device-tree, to be
later consumed by prom_init.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
[ bauerman: Minor adjustments to some comments. ]
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-4-bauerman@linux.ibm.com
Introduce CONFIG_PPC_SVM to control support for secure guests and include
Ultravisor-related helpers when it is selected
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820021326.6884-3-bauerman@linux.ibm.com
This branch has some cross-arch patches that are a prequisite for the
SVM work. They're in a topic branch in case any of the other arch
maintainers want to merge them to resolve conflicts.
The ultravisor (UV) provides an in-memory console which follows the
OPAL in-memory console structure.
This patch extends the OPAL msglog code to initialize the UV memory
console and provide the "/sys/firmware/ultravisor/msglog" interface
for userspace to view the UV message log.
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Tested-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190828130521.26764-2-mpe@ellerman.id.au
This patch refactors the code in opal-msglog that operates on the OPAL
memory console in order to make it cleaner and also allow the reuse of
the new memcons_* functions.
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Tested-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190828130521.26764-1-mpe@ellerman.id.au
When an SVM makes an hypercall or incurs some other exception, the
Ultravisor usually forwards (a.k.a. reflects) the exceptions to the
Hypervisor. After processing the exception, Hypervisor uses the
UV_RETURN ultracall to return control back to the SVM.
The expected register state on entry to this ultracall is:
* Non-volatile registers are restored to their original values.
* If returning from an hypercall, register R0 contains the return value
(unlike other ultracalls) and, registers R4 through R12 contain any
output values of the hypercall.
* R3 contains the ultracall number, i.e UV_RETURN.
* If returning with a synthesized interrupt, R2 contains the
synthesized interrupt number.
Thanks to input from Paul Mackerras, Ram Pai and Mike Anderson.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190822034838.27876-8-cclaudio@linux.ibm.com
LDBAR is a per-thread SPR populated and used by the thread-imc pmu
driver to dump the data counter into memory. It contains memory along
with few other configuration bits. LDBAR is populated and enabled only
when any of the thread imc pmu events are monitored.
In ultravisor enabled systems, LDBAR becomes ultravisor privileged and
an attempt to write to it will cause a Hypervisor Emulation Assistance
interrupt.
In ultravisor enabled systems, the ultravisor is responsible to maintain
the LDBAR (e.g. save and restore it).
This restricts LDBAR access to only when ultravisor is disabled.
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Reviewed-by: Ram Pai <linuxram@us.ibm.com>
Reviewed-by: Ryan Grimm <grimm@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190822034838.27876-7-cclaudio@linux.ibm.com
In ultravisor enabled systems, PTCR becomes ultravisor privileged only
for writing and an attempt to write to it will cause a Hypervisor
Emulation Assitance interrupt.
This patch uses the set_ptcr_when_no_uv() function to restrict PTCR
writing to only when ultravisor is disabled.
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190822034838.27876-6-cclaudio@linux.ibm.com
When Ultravisor (UV) is enabled, the partition table is stored in secure
memory and can only be accessed via the UV. The Hypervisor (HV) however
maintains a copy of the partition table in normal memory to allow Nest MMU
translations to occur (for normal VMs). The HV copy includes partition
table entries (PATE)s for secure VMs which would currently be unused
(Nest MMU translations cannot access secure memory) but they would be
needed as we add functionality.
This patch adds the UV_WRITE_PATE ucall which is used to update the PATE
for a VM (both normal and secure) when Ultravisor is enabled.
Signed-off-by: Michael Anderson <andmike@linux.ibm.com>
Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
[ cclaudio: Write the PATE in HV's table before doing that in UV's ]
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Reviewed-by: Ryan Grimm <grimm@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190822034838.27876-5-cclaudio@linux.ibm.com
In PEF enabled systems, some of the resources which were previously
hypervisor privileged are now ultravisor privileged and controlled by
the ultravisor firmware.
This adds FW_FEATURE_ULTRAVISOR to indicate if PEF is enabled.
The host kernel can use FW_FEATURE_ULTRAVISOR, for instance, to skip
accessing resources (e.g. PTCR and LDBAR) in case PEF is enabled.
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
[ andmike: Device node name to "ibm,ultravisor" ]
Signed-off-by: Michael Anderson <andmike@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190822034838.27876-4-cclaudio@linux.ibm.com
The ultracalls (ucalls for short) allow the Secure Virtual Machines
(SVM)s and hypervisor to request services from the ultravisor such as
accessing a register or memory region that can only be accessed when
running in ultravisor-privileged mode.
This patch adds the ucall_norets() ultravisor call handler.
The specific service needed from an ucall is specified in register
R3 (the first parameter to the ucall). Other parameters to the
ucall, if any, are specified in registers R4 through R12.
Return value of all ucalls is in register R3. Other output values
from the ucall, if any, are returned in registers R4 through R12.
Each ucall returns specific error codes, applicable in the context
of the ucall. However, like with the PowerPC Architecture Platform
Reference (PAPR), if no specific error code is defined for a particular
situation, then the ucall will fallback to an erroneous
parameter-position based code. i.e U_PARAMETER, U_P2, U_P3 etc depending
on the ucall parameter that may have caused the error.
Every host kernel (powernv) needs to be able to do ucalls in case it
ends up being run in a machine with ultravisor enabled. Otherwise, the
kernel may crash early in boot trying to access ultravisor resources,
for instance, trying to set the partition table entry 0. Secure guests
also need to be able to do ucalls and its kernel may not have
CONFIG_PPC_POWERNV=y. For that reason, the ucall.S file is placed under
arch/powerpc/kernel.
If ultravisor is not enabled, the ucalls will be redirected to the
hypervisor which must handle/fail the call.
Thanks to inputs from Ram Pai and Michael Anderson.
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190822034838.27876-3-cclaudio@linux.ibm.com
Protected Execution Facility (PEF) is an architectural change for
POWER 9 that enables Secure Virtual Machines (SVMs). When enabled,
PEF adds a new higher privileged mode, called Ultravisor mode, to POWER
architecture. Along with the new mode there is new firmware called the
Protected Execution Ultravisor (or Ultravisor for short).
POWER 9 DD2.3 chips (PVR=0x004e1203) or greater will be PEF-capable.
Attached documentation provides an overview of PEF and defines the API
for various interfaces that must be implemented in the Ultravisor
firmware as well as in the KVM Hypervisor.
Based on input from Mike Anderson, Thiago Bauermann, Claudio Carvalho,
Ben Herrenschmidt, Guerney Hunt, Paul Mackerras.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.ibm.com>
Signed-off-by: Ram Pai <linuxram@linux.ibm.com>
Signed-off-by: Guerney Hunt <gdhh@linux.ibm.com>
Reviewed-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Reviewed-by: Michael Anderson <andmike@linux.ibm.com>
Reviewed-by: Thiago Bauermann <bauerman@linux.ibm.com>
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190822034838.27876-2-cclaudio@linux.ibm.com
The ELF note documentation describes the types and descriptors to be
used with the PowerPC namespace.
Signed-off-by: Maxiwell S. Garcia <maxiwell@linux.ibm.com>
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829155021.2915-3-maxiwell@linux.ibm.com
Add the PowerPC name and the PPC_ELFNOTE_CAPABILITIES type in the
kernel binary ELF note. This type is a bitmap that can be used to
advertise kernel capabilities to userland.
This patch also defines PPCCAP_ULTRAVISOR_BIT as being the bit zero.
Suggested-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Claudio Carvalho <cclaudio@linux.ibm.com>
[ maxiwell: Define the 'PowerPC' type in the elfnote.h ]
Signed-off-by: Maxiwell S. Garcia <maxiwell@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829155021.2915-2-maxiwell@linux.ibm.com
As now we have xchg_no_kill/tce_kill, these are not used anymore so
remove them.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829085252.72370-6-aik@ozlabs.ru
This is the last implementation of iommu_table_ops::exchange() which
we are about to remove.
This implements xchg_no_kill() for pseries. Since it is paravirtual
platform, the hypervisor does TCE invalidations and we do not have
to deal with it here, hence no tce_kill() hook.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829085252.72370-5-aik@ozlabs.ru
Invalidating a TCE cache entry for each updated TCE is quite expensive.
This makes use of the new iommu_table_ops::xchg_no_kill()/tce_kill()
callbacks to bring down the time spent in mapping a huge guest DMA window.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829085252.72370-4-aik@ozlabs.ru
Invalidating a TCE cache entry for each updated TCE is quite expensive.
This makes use of the new iommu_table_ops::xchg_no_kill()/tce_kill()
callbacks to bring down the time spent in mapping a huge guest DMA window;
roughly 20s to 10s for each guest's 100GB of DMA space.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829085252.72370-3-aik@ozlabs.ru
At the moment updates in a TCE table are made by iommu_table_ops::exchange
which update one TCE and invalidates an entry in the PHB/NPU TCE cache
via set of registers called "TCE Kill" (hence the naming).
Writing a TCE is a simple xchg() but invalidating the TCE cache is
a relatively expensive OPAL call. Mapping a 100GB guest with PCI+NPU
passed through devices takes about 20s.
Thankfully we can do better. Since such big mappings happen at the boot
time and when memory is plugged/onlined (i.e. not often), these requests
come in 512 pages so we call call OPAL 512 times less which brings 20s
from the above to less than 10s. Also, since TCE caches can be flushed
entirely, calling OPAL for 512 TCEs helps skiboot [1] to decide whether
to flush the entire cache or not.
This implements 2 new iommu_table_ops callbacks:
- xchg_no_kill() to update a single TCE with no TCE invalidation;
- tce_kill() to invalidate multiple TCEs.
This uses the same xchg_no_kill() callback for IODA1/2.
This implements 2 new wrappers on top of the new callbacks similar to
the existing iommu_tce_xchg().
This does not use the new callbacks yet, the next patches will;
so this should not cause any behavioral change.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829085252.72370-2-aik@ozlabs.ru
H_PUT_TCE_INDIRECT handlers receive a page with up to 512 TCEs from
a guest. Although we verify correctness of TCEs before we do anything
with the existing tables, there is a small window when a check in
kvmppc_tce_validate might pass and right after that the guest alters
the page with TCEs which can cause early exit from the handler and
leave srcu_read_lock(&vcpu->kvm->srcu) (virtual mode) or lock_rmap(rmap)
(real mode) locked.
This fixes the bug by jumping to the common exit code with an appropriate
unlock.
Fixes: 121f80ba68 ("KVM: PPC: VFIO: Add in-kernel acceleration for VFIO")
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190826045520.92153-1-aik@ozlabs.ru
The existing code uses bunch of hardcoded values from the PCI Bus
Binding to IEEE Std 1275 spec; and it does so in quite non-obvious
way.
This defines fields from the cell#0 of the "reg" property of a PCI
device and uses them for parsing.
This should cause no behavioral change.
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
[mpe: Unsplit some 80/81 char lines, space the code with some newlines]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190829084417.71873-1-aik@ozlabs.ru
This switches to using common code for the DMA allocations, including
potential use of the CMA allocator if configured.
Switching to the generic code enables DMA allocations from atomic
context, which is required by the DMA API documentation, and also
adds various other minor features drivers start relying upon. It
also makes sure we have on tested code base for all architectures
that require uncached pte bits for coherent DMA allocations.
Another advantage is that consistent memory allocations now share
the general vmalloc pool instead of needing an explicit careout
from it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Christophe Leroy <christophe.leroy@c-s.fr> # tested on 8xx
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190814132230.31874-2-hch@lst.de
There is support for the kernel to execute the 'sc 0' instruction and
make a system call to itself. This is a relic that is unused in the
tree, therefore untested. It's also highly questionable for modules to
be doing this.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190827033010.28090-3-npiggin@gmail.com
Commit 3033f14ab7 ("clone: support passing tls argument via C rather
than pt_regs magic") introduced the HAVE_COPY_THREAD_TLS option. Use it
to avoid a subtle assumption about the argument ordering of clone type
syscalls.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190827033010.28090-2-npiggin@gmail.com
Nicholas Piggin
Ryan Grimm
Anshuman Khandual
Thiago Jung Bauermann
Sukadev Bhattiprolu
Ram Pai
Sukadev Bhattiprolu
Benjamin Herrenschmidt
Michael Ellerman
Claudio Carvalho
Michael Anderson
Maxiwell S. Garcia
Alexey Kardashevskiy
Christoph Hellwig
Christophe Leroy