xref: /rk3399_ARM-atf/docs/plat/arm/fvp/index.rst (revision f0bbaebc7ec00950908083e41352e856b65057f4)
1Arm Fixed Virtual Platforms (FVP)
2=================================
3
4Fixed Virtual Platform (FVP) Support
5------------------------------------
6
7This section lists the supported Arm |FVP| platforms. Please refer to the FVP
8documentation for a detailed description of the model parameter options.
9
10The latest version of the AArch64 build of TF-A has been tested on the following
11Arm FVPs without shifted affinities, and that do not support threaded CPU cores
12(64-bit host machine only).
13
14.. note::
15   The FVP models used are Version 11.9 Build 41, unless otherwise stated.
16
17-  ``FVP_Base_AEMv8A-AEMv8A``
18-  ``FVP_Base_AEMv8A-AEMv8A-AEMv8A-AEMv8A-CCN502``
19-  ``FVP_Base_RevC-2xAEMv8A``
20-  ``FVP_Base_Cortex-A32x4``
21-  ``FVP_Base_Cortex-A35x4``
22-  ``FVP_Base_Cortex-A53x4``
23-  ``FVP_Base_Cortex-A55x4+Cortex-A75x4``
24-  ``FVP_Base_Cortex-A55x4``
25-  ``FVP_Base_Cortex-A57x1-A53x1``
26-  ``FVP_Base_Cortex-A57x2-A53x4``
27-  ``FVP_Base_Cortex-A57x4-A53x4``
28-  ``FVP_Base_Cortex-A57x4``
29-  ``FVP_Base_Cortex-A65x4``
30-  ``FVP_Base_Cortex-A65AEx8``
31-  ``FVP_Base_Cortex-A72x4-A53x4``
32-  ``FVP_Base_Cortex-A72x4``
33-  ``FVP_Base_Cortex-A73x4-A53x4``
34-  ``FVP_Base_Cortex-A73x4``
35-  ``FVP_Base_Cortex-A75x4``
36-  ``FVP_Base_Cortex-A76x4``
37-  ``FVP_Base_Cortex-A76AEx4``
38-  ``FVP_Base_Cortex-A76AEx8``
39-  ``FVP_Base_Cortex-A77x4``
40-  ``FVP_Base_Neoverse-E1x1``
41-  ``FVP_Base_Neoverse-E1x2``
42-  ``FVP_Base_Neoverse-E1x4``
43-  ``FVP_Base_Neoverse-N1x4``
44-  ``FVP_Base_Zeusx4``
45-  ``FVP_CSS_SGI-575`` (Version 11.10 build 25)
46-  ``FVP_CSS_SGM-775``
47-  ``FVP_RD_E1Edge``
48-  ``FVP_RD_N1Edge`` (Version 11.10 build 25)
49-  ``Foundation_Platform``
50
51The latest version of the AArch32 build of TF-A has been tested on the
52following Arm FVPs without shifted affinities, and that do not support threaded
53CPU cores (64-bit host machine only).
54
55-  ``FVP_Base_AEMv8A-AEMv8A``
56-  ``FVP_Base_Cortex-A32x4``
57
58.. note::
59   The ``FVP_Base_RevC-2xAEMv8A`` FVP only supports shifted affinities, which
60   is not compatible with legacy GIC configurations. Therefore this FVP does not
61   support these legacy GIC configurations.
62
63The *Foundation* and *Base* FVPs can be downloaded free of charge. See the `Arm
64FVP website`_. The Cortex-A models listed above are also available to download
65from `Arm's website`_.
66
67.. note::
68   The build numbers quoted above are those reported by launching the FVP
69   with the ``--version`` parameter.
70
71.. note::
72   Linaro provides a ramdisk image in prebuilt FVP configurations and full
73   file systems that can be downloaded separately. To run an FVP with a virtio
74   file system image an additional FVP configuration option
75   ``-C bp.virtioblockdevice.image_path="<path-to>/<file-system-image>`` can be
76   used.
77
78.. note::
79   The software will not work on Version 1.0 of the Foundation FVP.
80   The commands below would report an ``unhandled argument`` error in this case.
81
82.. note::
83   FVPs can be launched with ``--cadi-server`` option such that a
84   CADI-compliant debugger (for example, Arm DS-5) can connect to and control
85   its execution.
86
87.. warning::
88   Since FVP model Version 11.0 Build 11.0.34 and Version 8.5 Build 0.8.5202
89   the internal synchronisation timings changed compared to older versions of
90   the models. The models can be launched with ``-Q 100`` option if they are
91   required to match the run time characteristics of the older versions.
92
93All the above platforms have been tested with `Linaro Release 19.06`_.
94
95.. _build_options_arm_fvp_platform:
96
97Arm FVP Platform Specific Build Options
98---------------------------------------
99
100-  ``FVP_CLUSTER_COUNT`` : Configures the cluster count to be used to
101   build the topology tree within TF-A. By default TF-A is configured for dual
102   cluster topology and this option can be used to override the default value.
103
104-  ``FVP_INTERCONNECT_DRIVER``: Selects the interconnect driver to be built. The
105   default interconnect driver depends on the value of ``FVP_CLUSTER_COUNT`` as
106   explained in the options below:
107
108   -  ``FVP_CCI`` : The CCI driver is selected. This is the default
109      if 0 < ``FVP_CLUSTER_COUNT`` <= 2.
110   -  ``FVP_CCN`` : The CCN driver is selected. This is the default
111      if ``FVP_CLUSTER_COUNT`` > 2.
112
113-  ``FVP_MAX_CPUS_PER_CLUSTER``: Sets the maximum number of CPUs implemented in
114   a single cluster.  This option defaults to 4.
115
116-  ``FVP_MAX_PE_PER_CPU``: Sets the maximum number of PEs implemented on any CPU
117   in the system. This option defaults to 1. Note that the build option
118   ``ARM_PLAT_MT`` doesn't have any effect on FVP platforms.
119
120-  ``FVP_USE_GIC_DRIVER`` : Selects the GIC driver to be built. Options:
121
122   -  ``FVP_GICV2`` : The GICv2 only driver is selected
123   -  ``FVP_GICV3`` : The GICv3 only driver is selected (default option)
124
125-  ``FVP_USE_SP804_TIMER`` : Use the SP804 timer instead of the Generic Timer
126   for functions that wait for an arbitrary time length (udelay and mdelay).
127   The default value is 0.
128
129-  ``FVP_HW_CONFIG_DTS`` : Specify the path to the DTS file to be compiled
130   to DTB and packaged in FIP as the HW_CONFIG. See :ref:`Firmware Design` for
131   details on HW_CONFIG. By default, this is initialized to a sensible DTS
132   file in ``fdts/`` folder depending on other build options. But some cases,
133   like shifted affinity format for MPIDR, cannot be detected at build time
134   and this option is needed to specify the appropriate DTS file.
135
136-  ``FVP_HW_CONFIG`` : Specify the path to the HW_CONFIG blob to be packaged in
137   FIP. See :ref:`Firmware Design` for details on HW_CONFIG. This option is
138   similar to the ``FVP_HW_CONFIG_DTS`` option, but it directly specifies the
139   HW_CONFIG blob instead of the DTS file. This option is useful to override
140   the default HW_CONFIG selected by the build system.
141
142Booting Firmware Update images
143------------------------------
144
145When Firmware Update (FWU) is enabled there are at least 2 new images
146that have to be loaded, the Non-Secure FWU ROM (NS-BL1U), and the
147FWU FIP.
148
149The additional fip images must be loaded with:
150
151::
152
153    --data cluster0.cpu0="<path_to>/ns_bl1u.bin"@0x0beb8000	[ns_bl1u_base_address]
154    --data cluster0.cpu0="<path_to>/fwu_fip.bin"@0x08400000	[ns_bl2u_base_address]
155
156The address ns_bl1u_base_address is the value of NS_BL1U_BASE.
157In the same way, the address ns_bl2u_base_address is the value of
158NS_BL2U_BASE.
159
160Booting an EL3 payload
161----------------------
162
163The EL3 payloads boot flow requires the CPU's mailbox to be cleared at reset for
164the secondary CPUs holding pen to work properly. Unfortunately, its reset value
165is undefined on the FVP platform and the FVP platform code doesn't clear it.
166Therefore, one must modify the way the model is normally invoked in order to
167clear the mailbox at start-up.
168
169One way to do that is to create an 8-byte file containing all zero bytes using
170the following command:
171
172.. code:: shell
173
174    dd if=/dev/zero of=mailbox.dat bs=1 count=8
175
176and pre-load it into the FVP memory at the mailbox address (i.e. ``0x04000000``)
177using the following model parameters:
178
179::
180
181    --data cluster0.cpu0=mailbox.dat@0x04000000   [Base FVPs]
182    --data=mailbox.dat@0x04000000                 [Foundation FVP]
183
184To provide the model with the EL3 payload image, the following methods may be
185used:
186
187#. If the EL3 payload is able to execute in place, it may be programmed into
188   flash memory. On Base Cortex and AEM FVPs, the following model parameter
189   loads it at the base address of the NOR FLASH1 (the NOR FLASH0 is already
190   used for the FIP):
191
192   ::
193
194       -C bp.flashloader1.fname="<path-to>/<el3-payload>"
195
196   On Foundation FVP, there is no flash loader component and the EL3 payload
197   may be programmed anywhere in flash using method 3 below.
198
199#. When using the ``SPIN_ON_BL1_EXIT=1`` loading method, the following DS-5
200   command may be used to load the EL3 payload ELF image over JTAG:
201
202   ::
203
204       load <path-to>/el3-payload.elf
205
206#. The EL3 payload may be pre-loaded in volatile memory using the following
207   model parameters:
208
209   ::
210
211       --data cluster0.cpu0="<path-to>/el3-payload>"@address   [Base FVPs]
212       --data="<path-to>/<el3-payload>"@address                [Foundation FVP]
213
214   The address provided to the FVP must match the ``EL3_PAYLOAD_BASE`` address
215   used when building TF-A.
216
217Booting a preloaded kernel image (Base FVP)
218-------------------------------------------
219
220The following example uses a simplified boot flow by directly jumping from the
221TF-A to the Linux kernel, which will use a ramdisk as filesystem. This can be
222useful if both the kernel and the device tree blob (DTB) are already present in
223memory (like in FVP).
224
225For example, if the kernel is loaded at ``0x80080000`` and the DTB is loaded at
226address ``0x82000000``, the firmware can be built like this:
227
228.. code:: shell
229
230    CROSS_COMPILE=aarch64-none-elf-  \
231    make PLAT=fvp DEBUG=1             \
232    RESET_TO_BL31=1                   \
233    ARM_LINUX_KERNEL_AS_BL33=1        \
234    PRELOADED_BL33_BASE=0x80080000    \
235    ARM_PRELOADED_DTB_BASE=0x82000000 \
236    all fip
237
238Now, it is needed to modify the DTB so that the kernel knows the address of the
239ramdisk. The following script generates a patched DTB from the provided one,
240assuming that the ramdisk is loaded at address ``0x84000000``. Note that this
241script assumes that the user is using a ramdisk image prepared for U-Boot, like
242the ones provided by Linaro. If using a ramdisk without this header,the ``0x40``
243offset in ``INITRD_START`` has to be removed.
244
245.. code:: bash
246
247    #!/bin/bash
248
249    # Path to the input DTB
250    KERNEL_DTB=<path-to>/<fdt>
251    # Path to the output DTB
252    PATCHED_KERNEL_DTB=<path-to>/<patched-fdt>
253    # Base address of the ramdisk
254    INITRD_BASE=0x84000000
255    # Path to the ramdisk
256    INITRD=<path-to>/<ramdisk.img>
257
258    # Skip uboot header (64 bytes)
259    INITRD_START=$(printf "0x%x" $((${INITRD_BASE} + 0x40)) )
260    INITRD_SIZE=$(stat -Lc %s ${INITRD})
261    INITRD_END=$(printf "0x%x" $((${INITRD_BASE} + ${INITRD_SIZE})) )
262
263    CHOSEN_NODE=$(echo                                        \
264    "/ {                                                      \
265            chosen {                                          \
266                    linux,initrd-start = <${INITRD_START}>;   \
267                    linux,initrd-end = <${INITRD_END}>;       \
268            };                                                \
269    };")
270
271    echo $(dtc -O dts -I dtb ${KERNEL_DTB}) ${CHOSEN_NODE} |  \
272            dtc -O dtb -o ${PATCHED_KERNEL_DTB} -
273
274And the FVP binary can be run with the following command:
275
276.. code:: shell
277
278    <path-to>/FVP_Base_AEMv8A-AEMv8A                            \
279    -C pctl.startup=0.0.0.0                                     \
280    -C bp.secure_memory=1                                       \
281    -C cluster0.NUM_CORES=4                                     \
282    -C cluster1.NUM_CORES=4                                     \
283    -C cache_state_modelled=1                                   \
284    -C cluster0.cpu0.RVBAR=0x04001000                           \
285    -C cluster0.cpu1.RVBAR=0x04001000                           \
286    -C cluster0.cpu2.RVBAR=0x04001000                           \
287    -C cluster0.cpu3.RVBAR=0x04001000                           \
288    -C cluster1.cpu0.RVBAR=0x04001000                           \
289    -C cluster1.cpu1.RVBAR=0x04001000                           \
290    -C cluster1.cpu2.RVBAR=0x04001000                           \
291    -C cluster1.cpu3.RVBAR=0x04001000                           \
292    --data cluster0.cpu0="<path-to>/bl31.bin"@0x04001000        \
293    --data cluster0.cpu0="<path-to>/<patched-fdt>"@0x82000000   \
294    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000 \
295    --data cluster0.cpu0="<path-to>/<ramdisk.img>"@0x84000000
296
297Obtaining the Flattened Device Trees
298^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
299
300Depending on the FVP configuration and Linux configuration used, different
301FDT files are required. FDT source files for the Foundation and Base FVPs can
302be found in the TF-A source directory under ``fdts/``. The Foundation FVP has
303a subset of the Base FVP components. For example, the Foundation FVP lacks
304CLCD and MMC support, and has only one CPU cluster.
305
306.. note::
307   It is not recommended to use the FDTs built along the kernel because not
308   all FDTs are available from there.
309
310The dynamic configuration capability is enabled in the firmware for FVPs.
311This means that the firmware can authenticate and load the FDT if present in
312FIP. A default FDT is packaged into FIP during the build based on
313the build configuration. This can be overridden by using the ``FVP_HW_CONFIG``
314or ``FVP_HW_CONFIG_DTS`` build options (refer to
315:ref:`build_options_arm_fvp_platform` for details on the options).
316
317-  ``fvp-base-gicv2-psci.dts``
318
319   For use with models such as the Cortex-A57-A53 Base FVPs without shifted
320   affinities and with Base memory map configuration.
321
322-  ``fvp-base-gicv2-psci-aarch32.dts``
323
324   For use with models such as the Cortex-A32 Base FVPs without shifted
325   affinities and running Linux in AArch32 state with Base memory map
326   configuration.
327
328-  ``fvp-base-gicv3-psci.dts``
329
330   For use with models such as the Cortex-A57-A53 Base FVPs without shifted
331   affinities and with Base memory map configuration and Linux GICv3 support.
332
333-  ``fvp-base-gicv3-psci-1t.dts``
334
335   For use with models such as the AEMv8-RevC Base FVP with shifted affinities,
336   single threaded CPUs, Base memory map configuration and Linux GICv3 support.
337
338-  ``fvp-base-gicv3-psci-dynamiq.dts``
339
340   For use with models as the Cortex-A55-A75 Base FVPs with shifted affinities,
341   single cluster, single threaded CPUs, Base memory map configuration and Linux
342   GICv3 support.
343
344-  ``fvp-base-gicv3-psci-aarch32.dts``
345
346   For use with models such as the Cortex-A32 Base FVPs without shifted
347   affinities and running Linux in AArch32 state with Base memory map
348   configuration and Linux GICv3 support.
349
350-  ``fvp-foundation-gicv2-psci.dts``
351
352   For use with Foundation FVP with Base memory map configuration.
353
354-  ``fvp-foundation-gicv3-psci.dts``
355
356   (Default) For use with Foundation FVP with Base memory map configuration
357   and Linux GICv3 support.
358
359
360Running on the Foundation FVP with reset to BL1 entrypoint
361^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
362
363The following ``Foundation_Platform`` parameters should be used to boot Linux with
3644 CPUs using the AArch64 build of TF-A.
365
366.. code:: shell
367
368    <path-to>/Foundation_Platform                   \
369    --cores=4                                       \
370    --arm-v8.0                                      \
371    --secure-memory                                 \
372    --visualization                                 \
373    --gicv3                                         \
374    --data="<path-to>/<bl1-binary>"@0x0             \
375    --data="<path-to>/<FIP-binary>"@0x08000000      \
376    --data="<path-to>/<kernel-binary>"@0x80080000   \
377    --data="<path-to>/<ramdisk-binary>"@0x84000000
378
379Notes:
380
381-  BL1 is loaded at the start of the Trusted ROM.
382-  The Firmware Image Package is loaded at the start of NOR FLASH0.
383-  The firmware loads the FDT packaged in FIP to the DRAM. The FDT load address
384   is specified via the ``hw_config_addr`` property in `TB_FW_CONFIG for FVP`_.
385-  The default use-case for the Foundation FVP is to use the ``--gicv3`` option
386   and enable the GICv3 device in the model. Note that without this option,
387   the Foundation FVP defaults to legacy (Versatile Express) memory map which
388   is not supported by TF-A.
389-  In order for TF-A to run correctly on the Foundation FVP, the architecture
390   versions must match. The Foundation FVP defaults to the highest v8.x
391   version it supports but the default build for TF-A is for v8.0. To avoid
392   issues either start the Foundation FVP to use v8.0 architecture using the
393   ``--arm-v8.0`` option, or build TF-A with an appropriate value for
394   ``ARM_ARCH_MINOR``.
395
396Running on the AEMv8 Base FVP with reset to BL1 entrypoint
397^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
398
399The following ``FVP_Base_RevC-2xAEMv8A`` parameters should be used to boot Linux
400with 8 CPUs using the AArch64 build of TF-A.
401
402.. code:: shell
403
404    <path-to>/FVP_Base_RevC-2xAEMv8A                            \
405    -C pctl.startup=0.0.0.0                                     \
406    -C bp.secure_memory=1                                       \
407    -C bp.tzc_400.diagnostics=1                                 \
408    -C cluster0.NUM_CORES=4                                     \
409    -C cluster1.NUM_CORES=4                                     \
410    -C cache_state_modelled=1                                   \
411    -C bp.secureflashloader.fname="<path-to>/<bl1-binary>"      \
412    -C bp.flashloader0.fname="<path-to>/<FIP-binary>"           \
413    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000 \
414    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
415
416.. note::
417   The ``FVP_Base_RevC-2xAEMv8A`` has shifted affinities and requires
418   a specific DTS for all the CPUs to be loaded.
419
420Running on the AEMv8 Base FVP (AArch32) with reset to BL1 entrypoint
421^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
422
423The following ``FVP_Base_AEMv8A-AEMv8A`` parameters should be used to boot Linux
424with 8 CPUs using the AArch32 build of TF-A.
425
426.. code:: shell
427
428    <path-to>/FVP_Base_AEMv8A-AEMv8A                            \
429    -C pctl.startup=0.0.0.0                                     \
430    -C bp.secure_memory=1                                       \
431    -C bp.tzc_400.diagnostics=1                                 \
432    -C cluster0.NUM_CORES=4                                     \
433    -C cluster1.NUM_CORES=4                                     \
434    -C cache_state_modelled=1                                   \
435    -C cluster0.cpu0.CONFIG64=0                                 \
436    -C cluster0.cpu1.CONFIG64=0                                 \
437    -C cluster0.cpu2.CONFIG64=0                                 \
438    -C cluster0.cpu3.CONFIG64=0                                 \
439    -C cluster1.cpu0.CONFIG64=0                                 \
440    -C cluster1.cpu1.CONFIG64=0                                 \
441    -C cluster1.cpu2.CONFIG64=0                                 \
442    -C cluster1.cpu3.CONFIG64=0                                 \
443    -C bp.secureflashloader.fname="<path-to>/<bl1-binary>"      \
444    -C bp.flashloader0.fname="<path-to>/<FIP-binary>"           \
445    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000 \
446    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
447
448Running on the Cortex-A57-A53 Base FVP with reset to BL1 entrypoint
449^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
450
451The following ``FVP_Base_Cortex-A57x4-A53x4`` model parameters should be used to
452boot Linux with 8 CPUs using the AArch64 build of TF-A.
453
454.. code:: shell
455
456    <path-to>/FVP_Base_Cortex-A57x4-A53x4                       \
457    -C pctl.startup=0.0.0.0                                     \
458    -C bp.secure_memory=1                                       \
459    -C bp.tzc_400.diagnostics=1                                 \
460    -C cache_state_modelled=1                                   \
461    -C bp.secureflashloader.fname="<path-to>/<bl1-binary>"      \
462    -C bp.flashloader0.fname="<path-to>/<FIP-binary>"           \
463    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000 \
464    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
465
466Running on the Cortex-A32 Base FVP (AArch32) with reset to BL1 entrypoint
467^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
468
469The following ``FVP_Base_Cortex-A32x4`` model parameters should be used to
470boot Linux with 4 CPUs using the AArch32 build of TF-A.
471
472.. code:: shell
473
474    <path-to>/FVP_Base_Cortex-A32x4                             \
475    -C pctl.startup=0.0.0.0                                     \
476    -C bp.secure_memory=1                                       \
477    -C bp.tzc_400.diagnostics=1                                 \
478    -C cache_state_modelled=1                                   \
479    -C bp.secureflashloader.fname="<path-to>/<bl1-binary>"      \
480    -C bp.flashloader0.fname="<path-to>/<FIP-binary>"           \
481    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000 \
482    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
483
484
485Running on the AEMv8 Base FVP with reset to BL31 entrypoint
486^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
487
488The following ``FVP_Base_RevC-2xAEMv8A`` parameters should be used to boot Linux
489with 8 CPUs using the AArch64 build of TF-A.
490
491.. code:: shell
492
493    <path-to>/FVP_Base_RevC-2xAEMv8A                             \
494    -C pctl.startup=0.0.0.0                                      \
495    -C bp.secure_memory=1                                        \
496    -C bp.tzc_400.diagnostics=1                                  \
497    -C cluster0.NUM_CORES=4                                      \
498    -C cluster1.NUM_CORES=4                                      \
499    -C cache_state_modelled=1                                    \
500    -C cluster0.cpu0.RVBAR=0x04010000                            \
501    -C cluster0.cpu1.RVBAR=0x04010000                            \
502    -C cluster0.cpu2.RVBAR=0x04010000                            \
503    -C cluster0.cpu3.RVBAR=0x04010000                            \
504    -C cluster1.cpu0.RVBAR=0x04010000                            \
505    -C cluster1.cpu1.RVBAR=0x04010000                            \
506    -C cluster1.cpu2.RVBAR=0x04010000                            \
507    -C cluster1.cpu3.RVBAR=0x04010000                            \
508    --data cluster0.cpu0="<path-to>/<bl31-binary>"@0x04010000    \
509    --data cluster0.cpu0="<path-to>/<bl32-binary>"@0xff000000    \
510    --data cluster0.cpu0="<path-to>/<bl33-binary>"@0x88000000    \
511    --data cluster0.cpu0="<path-to>/<fdt>"@0x82000000            \
512    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000  \
513    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
514
515Notes:
516
517-  If Position Independent Executable (PIE) support is enabled for BL31
518   in this config, it can be loaded at any valid address for execution.
519
520-  Since a FIP is not loaded when using BL31 as reset entrypoint, the
521   ``--data="<path-to><bl31|bl32|bl33-binary>"@<base-address-of-binary>``
522   parameter is needed to load the individual bootloader images in memory.
523   BL32 image is only needed if BL31 has been built to expect a Secure-EL1
524   Payload. For the same reason, the FDT needs to be compiled from the DT source
525   and loaded via the ``--data cluster0.cpu0="<path-to>/<fdt>"@0x82000000``
526   parameter.
527
528-  The ``FVP_Base_RevC-2xAEMv8A`` has shifted affinities and requires a
529   specific DTS for all the CPUs to be loaded.
530
531-  The ``-C cluster<X>.cpu<Y>.RVBAR=@<base-address-of-bl31>`` parameter, where
532   X and Y are the cluster and CPU numbers respectively, is used to set the
533   reset vector for each core.
534
535-  Changing the default value of ``ARM_TSP_RAM_LOCATION`` will also require
536   changing the value of
537   ``--data="<path-to><bl32-binary>"@<base-address-of-bl32>`` to the new value of
538   ``BL32_BASE``.
539
540
541Running on the AEMv8 Base FVP (AArch32) with reset to SP_MIN entrypoint
542^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
543
544The following ``FVP_Base_AEMv8A-AEMv8A`` parameters should be used to boot Linux
545with 8 CPUs using the AArch32 build of TF-A.
546
547.. code:: shell
548
549    <path-to>/FVP_Base_AEMv8A-AEMv8A                             \
550    -C pctl.startup=0.0.0.0                                      \
551    -C bp.secure_memory=1                                        \
552    -C bp.tzc_400.diagnostics=1                                  \
553    -C cluster0.NUM_CORES=4                                      \
554    -C cluster1.NUM_CORES=4                                      \
555    -C cache_state_modelled=1                                    \
556    -C cluster0.cpu0.CONFIG64=0                                  \
557    -C cluster0.cpu1.CONFIG64=0                                  \
558    -C cluster0.cpu2.CONFIG64=0                                  \
559    -C cluster0.cpu3.CONFIG64=0                                  \
560    -C cluster1.cpu0.CONFIG64=0                                  \
561    -C cluster1.cpu1.CONFIG64=0                                  \
562    -C cluster1.cpu2.CONFIG64=0                                  \
563    -C cluster1.cpu3.CONFIG64=0                                  \
564    -C cluster0.cpu0.RVBAR=0x04002000                            \
565    -C cluster0.cpu1.RVBAR=0x04002000                            \
566    -C cluster0.cpu2.RVBAR=0x04002000                            \
567    -C cluster0.cpu3.RVBAR=0x04002000                            \
568    -C cluster1.cpu0.RVBAR=0x04002000                            \
569    -C cluster1.cpu1.RVBAR=0x04002000                            \
570    -C cluster1.cpu2.RVBAR=0x04002000                            \
571    -C cluster1.cpu3.RVBAR=0x04002000                            \
572    --data cluster0.cpu0="<path-to>/<bl32-binary>"@0x04002000    \
573    --data cluster0.cpu0="<path-to>/<bl33-binary>"@0x88000000    \
574    --data cluster0.cpu0="<path-to>/<fdt>"@0x82000000            \
575    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000  \
576    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
577
578.. note::
579   The load address of ``<bl32-binary>`` depends on the value ``BL32_BASE``.
580   It should match the address programmed into the RVBAR register as well.
581
582Running on the Cortex-A57-A53 Base FVP with reset to BL31 entrypoint
583^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
584
585The following ``FVP_Base_Cortex-A57x4-A53x4`` model parameters should be used to
586boot Linux with 8 CPUs using the AArch64 build of TF-A.
587
588.. code:: shell
589
590    <path-to>/FVP_Base_Cortex-A57x4-A53x4                        \
591    -C pctl.startup=0.0.0.0                                      \
592    -C bp.secure_memory=1                                        \
593    -C bp.tzc_400.diagnostics=1                                  \
594    -C cache_state_modelled=1                                    \
595    -C cluster0.cpu0.RVBARADDR=0x04010000                        \
596    -C cluster0.cpu1.RVBARADDR=0x04010000                        \
597    -C cluster0.cpu2.RVBARADDR=0x04010000                        \
598    -C cluster0.cpu3.RVBARADDR=0x04010000                        \
599    -C cluster1.cpu0.RVBARADDR=0x04010000                        \
600    -C cluster1.cpu1.RVBARADDR=0x04010000                        \
601    -C cluster1.cpu2.RVBARADDR=0x04010000                        \
602    -C cluster1.cpu3.RVBARADDR=0x04010000                        \
603    --data cluster0.cpu0="<path-to>/<bl31-binary>"@0x04010000    \
604    --data cluster0.cpu0="<path-to>/<bl32-binary>"@0xff000000    \
605    --data cluster0.cpu0="<path-to>/<bl33-binary>"@0x88000000    \
606    --data cluster0.cpu0="<path-to>/<fdt>"@0x82000000            \
607    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000  \
608    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
609
610Running on the Cortex-A32 Base FVP (AArch32) with reset to SP_MIN entrypoint
611^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
612
613The following ``FVP_Base_Cortex-A32x4`` model parameters should be used to
614boot Linux with 4 CPUs using the AArch32 build of TF-A.
615
616.. code:: shell
617
618    <path-to>/FVP_Base_Cortex-A32x4                             \
619    -C pctl.startup=0.0.0.0                                     \
620    -C bp.secure_memory=1                                       \
621    -C bp.tzc_400.diagnostics=1                                 \
622    -C cache_state_modelled=1                                   \
623    -C cluster0.cpu0.RVBARADDR=0x04002000                       \
624    -C cluster0.cpu1.RVBARADDR=0x04002000                       \
625    -C cluster0.cpu2.RVBARADDR=0x04002000                       \
626    -C cluster0.cpu3.RVBARADDR=0x04002000                       \
627    --data cluster0.cpu0="<path-to>/<bl32-binary>"@0x04002000   \
628    --data cluster0.cpu0="<path-to>/<bl33-binary>"@0x88000000   \
629    --data cluster0.cpu0="<path-to>/<fdt>"@0x82000000           \
630    --data cluster0.cpu0="<path-to>/<kernel-binary>"@0x80080000 \
631    --data cluster0.cpu0="<path-to>/<ramdisk>"@0x84000000
632
633--------------
634
635*Copyright (c) 2019-2020, Arm Limited. All rights reserved.*
636
637.. _TB_FW_CONFIG for FVP: ../plat/arm/board/fvp/fdts/fvp_tb_fw_config.dts
638.. _Arm's website: `FVP models`_
639.. _FVP models: https://developer.arm.com/products/system-design/fixed-virtual-platforms
640.. _Linaro Release 19.06: http://releases.linaro.org/members/arm/platforms/19.06
641.. _Arm FVP website: https://developer.arm.com/products/system-design/fixed-virtual-platforms
642