This is useful, especially when testing relatively new gdbstub
features that might not be in distro packages yet.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20200316172155.971-22-alex.bennee@linaro.org>

This tests a bunch of registers that the kernel allows userspace to
read including the CPUID registers. We need a SVE aware compiler as we
are testing the id_aa64zfr0_el1 register in the set.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20200316172155.971-21-alex.bennee@linaro.org>

For system emulation we need to check the state of the GIC before we
report the value. However this isn't relevant to exporting of the
value to linux-user and indeed breaks the exported value as set by
modify_arm_cp_regs.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200316172155.971-20-alex.bennee@linaro.org>

We also expose a the helpers to read/write the the registers.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Richard Henderson <richard.henderson@linaro.org>

Message-Id: <20200316172155.971-19-alex.bennee@linaro.org>

The Linux kernel chooses the default of 64 bytes for SVE registers on
the basis that it is the largest size on known hardware that won't
grow the signal frame. We still honour the sve-max-vq property and
userspace can expand the number of lanes by calling PR_SVE_SET_VL.

This should not make any difference to SVE enabled software as the SVE
is of course vector length agnostic.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

Message-Id: <20200316172155.971-18-alex.bennee@linaro.org>

This is described as optional but I'm not convinced of the numbering
when multiple target fragments are sent.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

Message-Id: <20200316172155.971-17-alex.bennee@linaro.org>

We will want to generate similar dynamic XML for gdbstub support of
SVE registers (the upstream doesn't use XML). To that end lightly
rename a few things to make the distinction.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20200316172155.971-16-alex.bennee@linaro.org>

Instead of passing a pointer to memory now just extend the GByteArray
to all the read register helpers. They can then safely append their
data through the normal way. We don't bother with this abstraction for
write registers as we have already ensured the buffer being copied
from is the correct size.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Damien Hedde <damien.hedde@greensocs.com>

Message-Id: <20200316172155.971-15-alex.bennee@linaro.org>

This is cleaner than poking memory directly and will make later
clean-ups easier.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200316172155.971-14-alex.bennee@linaro.org>

This is cleaner than poking memory directly and will make later
clean-ups easier.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200316172155.971-13-alex.bennee@linaro.org>

This is cleaner than poking memory directly and will make later
clean-ups easier.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

Message-Id: <20200316172155.971-12-alex.bennee@linaro.org>

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20200316172155.971-11-alex.bennee@linaro.org>

This is in preparation for further re-factoring of the register API
with the rest of the code. Theoretically the read register function
could overwrite the MAX_PACKET_LENGTH buffer although currently all
registers are well within the size range.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Damien Hedde <damien.hedde@greensocs.com>
Tested-by: Damien Hedde <damien.hedde@greensocs.com>

Message-Id: <20200316172155.971-10-alex.bennee@linaro.org>

Rather than having a static buffer replace str_buf with a GString
which we know can grow on demand. Convert the internal functions to
take a GString instead of a char * and length.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Damien Hedde <damien.hedde@greensocs.com>
Tested-by: Damien Hedde <damien.hedde@greensocs.com>
Message-Id: <20200316172155.971-9-alex.bennee@linaro.org>

We only have one GDBState which should be allocated at the time we
process any commands. This will make further clean-up a bit easier.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Damien Hedde <damien.hedde@greensocs.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>

Message-Id: <20200316172155.971-8-alex.bennee@linaro.org>

Instead of allocating make this entirely static. We shall reduce the
size of the structure in later commits and dynamically allocate parts
of it. We introduce an init and reset helper function to keep all the
manipulation in one place.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Damien Hedde <damien.hedde@greensocs.com>

Message-Id: <20200316172155.971-7-alex.bennee@linaro.org>

Since commit f5852efa we can display GLib errors with the QEMU
error reporting API. Set it to the 'error' level, as this helps
understanding failures from QEMU calls to GLib on Travis-CI.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200316101544.22361-1-philmd@redhat.com>
Message-Id: <20200316172155.971-6-alex.bennee@linaro.org>

Building the qemu:debian-amd64 fails when building VirGL:

  make[2]: Entering directory '/usr/src/virglrenderer/src/gallium/auxiliary'
    CC       cso_cache/cso_cache.lo
    CC       cso_cache/cso_hash.lo
    CC       os/os_misc.lo
    CC       util/u_debug.lo
    CC       util/u_debug_describe.lo
    CC       util/u_format.lo
    GEN      util/u_format_table.c
  Traceback (most recent call last):
    File "./util/u_format_table.py", line 168, in <module>
      main()
    File "./util/u_format_table.py", line 164, in main
      write_format_table(formats)
    File "./util/u_format_table.py", line 132, in write_format_table
      print("   %s,\t/* is_array */" % (bool_map(format.is_array()),))
    File "/usr/src/virglrenderer/src/gallium/auxiliary/util/u_format_parse.py", line 164, in is_array
      return self.array_element() != None
    File "/usr/src/virglrenderer/src/gallium/auxiliary/util/u_format_parse.py", line 73, in __eq__
      return self.type == other.type and self.norm == other.norm and self.pure == other.pure and self.size == other.size
  AttributeError: 'NoneType' object has no attribute 'type'
  make[2]: Leaving directory '/usr/src/virglrenderer/src/gallium/auxiliary'
  make[2]: *** [Makefile:906: util/u_format_table.c] Error 1
  make[1]: *** [Makefile:631: install-recursive] Error 1

VirGL commits a8962eda1..a613dcc82 fix this problem.
Update to VirGL 0.8.0 which contains them.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200212202709.12665-4-philmd@redhat.com>
Message-Id: <20200316172155.971-5-alex.bennee@linaro.org>

The GLX configure option has been removed in 71c75f201d [*].
We missed that when updating to v0.7.0 in commit fab3220f.

This silents:

  configure: creating ./config.status
  config.status: creating virglrenderer.pc
  ...
  configure: WARNING: unrecognized options: --with-glx

[*] https://gitlab.freedesktop.org/virgl/virglrenderer/commit/71c75f201d



Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200212202709.12665-3-philmd@redhat.com>
Message-Id: <20200316172155.971-4-alex.bennee@linaro.org>

freedesktop.org is moving to a GitLab instance,
use the new url.

- https://www.fooishbar.org/blog/gitlab-fdo-introduction/
- https://gitlab.freedesktop.org/freedesktop/freedesktop/-/wikis/home



Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200212202709.12665-2-philmd@redhat.com>
Message-Id: <20200316172155.971-3-alex.bennee@linaro.org>

We often run Linux kernels to test QEMU. We sometimes need
to build them manually to use non-default features. We only
miss the tiny 'bc' tool.

The ncurses library is helpful to run 'make menuconfig'.

Finally, gdb-multiarch allow us to debug a TCG guest when its
architecture is different than the host.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200212202738.12986-1-philmd@redhat.com>
Message-Id: <20200316172155.971-2-alex.bennee@linaro.org>

target-arm:
 * hw/arm/pxa2xx: Do not wire up OHCI for PXA255
 * aspeed/smc: Fix number of dummy cycles for FAST_READ_4 command
 * m25p80: Improve command handling for Jedec and unsupported commands
 * hw/net/imx_fec: write TGSR and TCSR3 in imx_enet_write()
 * hw/arm/fsl-imx6, imx6ul: Wire up USB controllers
 * hw/arm/fsl-imx6ul: Instantiate unimplemented pwm and can devices

# gpg: Signature made Tue 17 Mar 2020 11:40:01 GMT
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [ultimate]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* remotes/pmaydell/tags/pull-target-arm-20200317:
  hw/arm/pxa2xx: Do not wire up OHCI for PXA255
  aspeed/smc: Fix number of dummy cycles for FAST_READ_4 command
  m25p80: Improve command handling for unsupported commands
  m25p80: Improve command handling for Jedec commands
  m25p80: Convert to support tracing
  hw/net/imx_fec: write TGSR and TCSR3 in imx_enet_write()
  hw/arm/fsl-imx6: Wire up USB controllers
  hw/arm/fsl-imx6ul: Wire up USB controllers
  hw/arm/fsl-imx6ul: Instantiate unimplemented pwm and can devices
  hw/arm/fsl-imx6ul: Fix USB interrupt numbers
  hw/usb: Add basic i.MX USB Phy support

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

usb: bugfixes for usb-serial @ xhci.

# gpg: Signature made Tue 17 Mar 2020 09:50:04 GMT
# gpg:                using RSA key 4CB6D8EED3E87138
# gpg: Good signature from "Gerd Hoffmann (work) <kraxel@redhat.com>" [full]
# gpg:                 aka "Gerd Hoffmann <gerd@kraxel.org>" [full]
# gpg:                 aka "Gerd Hoffmann (private) <kraxel@gmail.com>" [full]
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/usb-20200317-pull-request:
  usb-serial: Fix timeout closing the device
  usb-serial: Increase receive buffer to 496
  usb-serial: chunk data to wMaxPacketSize
  usb-serial: Move USB_TOKEN_IN into a helper function

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

PXA255 does not support a USB OHCI controller, so don't wire it up.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Message-id: 20200313160215.28155-1-linux@roeck-us.net
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

The Linux kernel recently started using FAST_READ_4 commands.
This results in flash read failures. At the same time, the m25p80
emulation is seen to read 8 more bytes than expected. Adjusting the
expected number of dummy cycles to match FAST_READ fixes the problem.

Fixes: f95c4bff ("aspeed/smc: snoop SPI transfers to fake dummy cycles")
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Whenever an unsupported command is encountered, the current code
interprets each transferred byte as new command. Most of the time, those
'commands' are interpreted as new unknown commands. However, in rare
cases, it may be that for example address or length information
passed with the original command is by itself a valid command.
If that happens, the state machine may get completely confused and,
worst case, start writing data into the flash or even erase it.

To avoid the problem, transition into STATE_READING_DATA and keep
sending a value of 0 until the chip is deselected after encountering
an unsupported command.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

When requesting JEDEC data using the JEDEC_READ command, the Linux kernel
always requests 6 bytes. The current implementation only returns three
bytes, and interprets the remaining three bytes as new commands.
While this does not matter most of the time, it is at the very least
confusing. To avoid the problem, always report up to 6 bytes of JEDEC
data. Fill remaining data with 0.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

While at it, add some trace messages to help debug problems
seen when running the latest Linux kernel.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

The current code causes clang static code analyzer generate warning:
hw/net/imx_fec.c:858:9: warning: Value stored to 'value' is never read
        value = value & 0x0000000f;
        ^       ~~~~~~~~~~~~~~~~~~
hw/net/imx_fec.c:864:9: warning: Value stored to 'value' is never read
        value = value & 0x000000fd;
        ^       ~~~~~~~~~~~~~~~~~~

According to the definition of the function, the two “value” assignments
 should be written to registers.

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Chen Qun <kuhn.chenqun@huawei.com>
Message-id: 20200313123242.13236-1-kuhn.chenqun@huawei.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

With this patch, the USB controllers on 'sabrelite' are detected
and can be used to boot the system.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Message-id: 20200313014551.12554-6-linux@roeck-us.net
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

IMX6UL USB controllers are quite similar to IMX7 USB controllers.
Wire them up the same way.

The only real difference is that wiring up phy devices is necessary
to avoid phy reset timeouts in the Linux kernel.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Message-id: 20200313014551.12554-5-linux@roeck-us.net
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Recent Linux kernels (post v4.20) crash due to accesses to flexcan
and pwm controllers. Instantiate as unimplemented devices to work
around the problem.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Message-id: 20200313014551.12554-4-linux@roeck-us.net
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

USB1 and USB2 interrupt numbers were swapped. USB_PHY2 interrupt number
is 45. That didn't really matter up to now since the interrupts were not
used, but it needs to be fixed to be able to wire up the USB controllers.

Fixes: 31cbf933 ("i.MX6UL: Add i.MX6UL SOC")
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Message-id: 20200313014551.12554-3-linux@roeck-us.net
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Add basic USB PHY support as implemented in i.MX23, i.MX28, i.MX6,
and i.MX7 SoCs.

The only support really needed - at least to boot Linux - is support
for soft reset, which needs to reset various registers to their initial
value. Otherwise, just record register values.

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Message-id: 20200313014551.12554-2-linux@roeck-us.net
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

RISC-V Patches for the 5.0 Soft Freeze, Part 5

This tag contains the last of the patches I'd like to target for the 5.0 soft
freeze.  At this point we're mostly collecting fixes, but there are a few new
features.  The changes include:

* An OpenSBI update, including the various bits necessary to put CI together
  and an image for the 32-bit sifive_u board.
* A fix that disallows TSR when outside of machine mode.
* A fix for VS-mode interrupt forwarding.

# gpg: Signature made Tue 17 Mar 2020 03:59:58 GMT
# gpg:                using RSA key 2B3C3747446843B24A943A7A2E1319F35FBB1889
# gpg:                issuer "palmer@dabbelt.com"
# gpg: Good signature from "Palmer Dabbelt <palmer@dabbelt.com>" [unknown]
# gpg:                 aka "Palmer Dabbelt <palmer@sifive.com>" [unknown]
# gpg:                 aka "Palmer Dabbelt <palmerdabbelt@google.com>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 00CE 76D1 8349 60DF CE88  6DF8 EF4C A150 2CCB AB41
#      Subkey fingerprint: 2B3C 3747 4468 43B2 4A94  3A7A 2E13 19F3 5FBB 1889

* remotes/palmer/tags/riscv-for-master-5.0-sf5:
  target/riscv: Fix VS mode interrupts forwarding.
  gitlab-ci.yml: Add jobs to build OpenSBI firmware binaries
  riscv: sifive_u: Update BIOS_FILENAME for 32-bit
  roms: opensbi: Add 32-bit firmware image for sifive_u machine
  roms: opensbi: Upgrade from v0.5 to v0.6
  target/riscv: Correctly implement TSR trap

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Linux guests wait ~30 seconds when closing the emulated /dev/ttyUSB0.
During that time, the kernel driver is sending many control URBs
requesting GetModemStat (5).  Real hardware returns a status with
FTDI_THRE (Transmitter Holding Register) and FTDI_TEMT (Transmitter
Empty) set.  QEMU leaves them clear, and it seems Linux is waiting for
FTDI_TEMT to be set to indicate the tx queue is empty before closing.

Set the bits when responding to a GetModemStat query and avoid the
shutdown delay.

Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
Reviewed-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Message-id: 20200316174610.115820-5-jandryuk@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

A FTDI USB adapter on an xHCI controller can send 512 byte USB packets.
These are 8 * ( 2 bytes header + 62 bytes data).  A 384 byte receive
buffer is insufficient to fill a 512 byte packet, so bump the receive
size to 496 ( 512 - 2 * 8 ).

Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
Reviewed-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Message-id: 20200316174610.115820-4-jandryuk@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

usb-serial has issues with xHCI controllers where data is lost in the
VM.  Inspecting the URBs in the guest, EHCI starts every 64 byte boundary
(wMaxPacketSize) with a header.  EHCI hands packets into
usb_serial_token_in() with size 64, so these cannot cross the 64 byte
boundary.  The xHCI controller has packets of 512 bytes and the usb-serial
will just write through the 64 byte boundary.  In the guest, this means
data bytes are interpreted as header, so data bytes don't make it out
the serial interface.

Re-work usb_serial_token_in to chunk data into 64 byte units - 2 byte
header and 62 bytes data.  The Linux driver reads wMaxPacketSize to find
the chunk size, so we match that.

Real hardware was observed to pass in 512 byte URBs (496 bytes data +
8 * 2 byte headers).  Since usb-serial only buffers 384 bytes of data,
usb-serial will pass in 6 64 byte blocks and 1 12 byte partial block for
462 bytes max.

Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
Message-id: 20200316174610.115820-3-jandryuk@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

We'll be adding a loop, so move the code into a helper function.  breaks
are replaced with returns.  While making this change, add braces to
single line if statements to comply with coding style and keep
checkpatch happy.

Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
Message-id: 20200316174610.115820-2-jandryuk@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

Currently riscv_cpu_local_irq_pending is used to find out pending
interrupt and VS mode interrupts are being shifted to represent
S mode interrupts in this function. So when the cause returned by
this function is passed to riscv_cpu_do_interrupt to actually
forward the interrupt, the VS mode forwarding check does not work
as intended and interrupt is actually forwarded to hypervisor. This
patch fixes this issue.

Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal49@gmail.com>
Reviewed-by: Palmer Dabbelt <palmerdabbelt@google.com>
Signed-off-by: Palmer Dabbelt <palmerdabbelt@google.com>