Commit graph

344 commits

Author SHA1 Message Date
Eduardo Habkost 074865ff98
cpu: Generify CPU init functions
Backports commits 2994fd96d986578a342f2342501b4ad30f6d0a85,
701e3c78ce45fa630ffc6826c4b9a4218954bc7f, and
d1853231c60d16af78cf4d1608d043614bfbac0b from qemuu
2018-03-20 08:21:51 -04:00
Peter Crosthwaite ce1831bfb4
target-*: Don't redefine cpu_exec()
This function needs to be converted to QOM hook and virtualised for
multi-arch. This rename interferes, as cpu-qom will not have access
to the renaming causing name divergence. This rename doesn't really do
anything anyway so just delete it.

Backports commit 8642c1b81e0418df066a7960a7426d85a923a253 from qemu
2018-03-20 07:02:47 -04:00
Igor Mammedov 15eb359656
numa: move numa_node from CPUState into target specific classes
Move vcpu's associated numa_node field out of generic CPUState
into inherited classes that actually care about cpu<->numa mapping,
i.e: ARMCPU, PowerPCCPU, X86CPU.

Backports relevant parts of commit 15f8b14228b856850df3fa5ba999ad96521f2208 from qemu
2018-03-20 06:46:20 -04:00
Lioncash 750d56421c
translate/arm/vec_helper: Align to qemu formatting 2018-03-12 11:59:14 -04:00
Lioncash bab31a2510
target/arm/cpu and crypto_helper: Correct bad merge and adjust to qemu code style 2018-03-12 11:57:24 -04:00
Lioncash 0751366e5c
target/arm/op_helper: Correct bad merge 2018-03-12 11:42:43 -04:00
Lioncash 9a0632bfcf
target/arm/helper64: Correct bad merge 2018-03-12 11:37:27 -04:00
Lioncash c93c3bd4b3
target/arm/helper: Correct bad merge 2018-03-12 11:33:45 -04:00
Lioncash 14c1fcd5bf
target/arm/translate: Correct bad merge 2018-03-12 11:17:37 -04:00
Lioncash 0dd13de42f
target/arm/translate-a64: Correct bad merge 2018-03-12 11:17:33 -04:00
Peter Maydell fabd6c7ae8
target/arm: Make 'any' CPU just an alias for 'max'
Now we have a working '-cpu max', the linux-user-only
'any' CPU is pretty much the same thing, so implement it
that way.

For the moment we don't add any of the extra feature bits
to the system-emulation "max", because we don't set the
ID register bits we would need to to advertise those
features as present.

Backports commit a0032cc5427d0d396aa0a9383ad9980533448ea4 from qemu
2018-03-12 10:11:49 -04:00
Peter Maydell 7388fff079
target/arm: Add "-cpu max" support
Add support for "-cpu max" for ARM guests. This CPU type behaves
like "-cpu host" when KVM is enabled, and like a system CPU with
the maximum possible feature set otherwise. (Note that this means
it won't be migratable across versions, as we will likely add
features to it in future.)

Backports commit bab52d4bba3f22921a690a887b4bd0342f2754cd from qemu
2018-03-12 10:11:49 -04:00
Alistair Francis 44d8c38138
target/arm: Add a core count property
The cortex A53 TRM specifies that bits 24 and 25 of the L2CTLR register
specify the number of cores in the processor, not the total number of
cores in the system. To report this correctly on machines with multiple
CPU clusters (ARM's big.LITTLE or Xilinx's ZynqMP) we need to allow
the machine to overwrite this value. To do this let's add an optional
property.

Backports commit f9a697112ee64180354f98309a5d6b691cc8699d from qemu
2018-03-12 10:11:48 -04:00
Lioncash 8e161bb723
target/arm: Use the any cpu model instead of cortex-a57
The Cortex-A57 doesn't allow use of v8.1+ architecture instructions
2018-03-12 03:42:57 -04:00
Eduardo Habkost a7f59d7771
Use DEFINE_MACHINE() to register all machines
Convert all machines to use DEFINE_MACHINE() instead of QEMUMachine
automatically using a script.

Backports commit e264d29de28c5b0be3d063307ce9fb613b427cc3 from qemu
2018-03-11 15:12:46 -04:00
Richard Henderson 81ae246f07
target/arm: Enable ARM_FEATURE_V8_FCMA
Enable it for the "any" CPU used by *-linux-user.

Backports commit e66a67bf28e1b4fce2e3d72a2610dbd48d9d3078 from qemu
2018-03-09 01:12:19 -05:00
Richard Henderson 85cfb78ea2
target/arm: Decode t32 simd 3reg and 2reg_scalar extension
Happily, the bits are in the same places compared to a32.

Backports commit 0052087efb8a5c0e29ddc2f59f8476fcdc6495b2 from qemu
2018-03-09 01:11:14 -05:00
Richard Henderson e5da25aaf8
target/arm: Decode aa32 armv8.3 2-reg-index
Backports commit 638808ff8a0c0d62333822d3756e5d98f9f369c3 from qemu
2018-03-09 01:09:59 -05:00
Richard Henderson 69890ae145
target/arm: Decode aa32 armv8.3 3-same
Backports commit 8b7209fae730813d722b17a8a13b6a16c84616c8 from qemu
2018-03-09 01:08:26 -05:00
Richard Henderson abd86b2287
target/arm: Decode aa64 armv8.3 fcmla
Backports commit d17b7cdcf4ea3e858ceee8b86fc8544bb71561e6 from qemu

Also remember to commit vec_helper.
2018-03-09 01:05:02 -05:00
Richard Henderson 4b39a36416
target/arm: Decode aa64 armv8.3 fcadd
Backports commit 1695cd61b08d4376c11e0658836c4f08b4fc3aa1 from qemu
2018-03-09 00:58:37 -05:00
Richard Henderson 0b1ab3e745
target/arm: Add ARM_FEATURE_V8_FCMA
Not enabled anywhere yet.

Backports commit 0438f0372a7031debe796f4e3d30875d4d1e7899 from qemu
2018-03-09 00:28:37 -05:00
Richard Henderson fc74a022bf
target/arm: Enable ARM_FEATURE_V8_RDM
Enable it for the "any" CPU used by *-linux-user.

Backports commit f5dfc2ecdd48b71900bc50298ad2768d60356e44 from qemu
2018-03-09 00:27:34 -05:00
Richard Henderson 78b0b9c523
target/arm: Decode aa32 armv8.1 two reg and a scalar
Backports commit 61adacc8f589539ac6b25cfcbd6e099357188974 from qemu
2018-03-09 00:24:14 -05:00
Richard Henderson ca4ceb2dd7
target/arm: Decode aa32 armv8.1 three same
Backports commit 36a719348a9744d17c6ef6bac01bcb5fcd279753 from qemu
2018-03-09 00:18:31 -05:00
Richard Henderson 152c9484bd
target/arm: Decode aa64 armv8.1 scalar/vector x indexed element
Backports commit d345df7a3f1336ceb0537c1fa0a7261030426768 from qemu
2018-03-09 00:12:00 -05:00
Lioncash 12fd2cc113
target/arm: Decode aa64 armv8.1 three same extra 2018-03-09 00:10:09 -05:00
Richard Henderson 4f585f71fb
target/arm: Decode aa64 armv8.1 scalar three same extra
Backports commit d9061ec3d27eb940402a7eafee3fb77ce1146ad4 from qemu
2018-03-09 00:02:23 -05:00
Richard Henderson 774cbded7a
target/arm: Refactor disas_simd_indexed size checks
The integer size check was already outside of the opcode switch;
move the floating-point size check outside as well. Unify the
size vs index adjustment between fp and integer paths.

Backports commit 449f264b1749ac0e59c58bbc2eacdb3dc302c2bf from qemu
2018-03-08 23:53:39 -05:00
Richard Henderson 1fd2644738
target/arm: Refactor disas_simd_indexed decode
Include the U bit in the switches rather than testing separately.

Backports commit 5f81b1de43259ed0969e62a7419ab9dd9da2c5c0 from qemu
2018-03-08 23:44:03 -05:00
Richard Henderson 109a777fd6
target/arm: Add ARM_FEATURE_V8_RDM
Not enabled anywhere yet.

Backports commit 1dc81c15418d9b174f59a1c6262eb3487f352c56 from qemu
2018-03-08 23:44:03 -05:00
Peter Maydell e917a1ac0e
target/arm: Add Cortex-M33
Add a Cortex-M33 definition. The M33 is an M profile CPU
which implements the ARM v8M architecture, including the
M profile Security Extension.

Backports commit c7b26382fee8b745c6e903c85281babf30c2cb7c from qemu
2018-03-08 23:44:03 -05:00
Peter Maydell bd606401dc
target/arm: Define init-svtor property for the reset secure VTOR value
The Cortex-M33 allows the system to specify the reset value of the
secure Vector Table Offset Register (VTOR) by asserting config
signals. In particular, guest images for the MPS2 AN505 board rely
on the MPS2's initial VTOR being correct for that board.
Implement a QEMU property so board and SoC code can set the reset
value to the correct value.

Backports commit 38e2a77c9d6876e58f45cabb1dd9a6a60c22b39e from qemu
2018-03-08 23:44:03 -05:00
Peter Maydell eb4796e965
target/arm: Enable ARM_V8_FP16 feature bit for the AArch64 any CPU
Now we have implemented FP16 we can enable it for the "any" CPU.

Backports commit 969b389ee8ba84bc3f2e7ccfa993679fac410ad2 from qemu
2018-03-08 23:44:02 -05:00
Alex Bennée 6e41113897
arm/translate-a64: add all single op FP16 to handle_fp_1src_half
This includes FMOV, FABS, FNEG, FSQRT and FRINT[NPMZAXI]. We re-use
existing helpers to achieve this.

Backports commit c2c08713a6a5846bbe601d4d1b4f9708ba77efdc from qemu
2018-03-08 23:44:02 -05:00
Alex Bennée c6c8a1cccc
arm/translate-a64: implement simd_scalar_three_reg_same_fp16
This covers the encoding group:

Advanced SIMD scalar three same FP16

As all the helpers are already there it is simply a case of calling the
existing helpers in the scalar context.

Backports commit 7c93b7741b29b3ffda81a6e9525771b4409db99f from qemu
2018-03-08 23:44:02 -05:00
Alex Bennée dd29452046
arm/translate-a64: add all FP16 ops in simd_scalar_pairwise
I only needed to do a little light re-factoring to support the
half-precision helpers.

Backports commit 5c36d89567cfd049a7c59ff219639f788225068f from qemu
2018-03-08 23:44:02 -05:00
Alex Bennée 8bbabd7eb3
arm/translate-a64: add FP16 FMOV to simd_mod_imm
Only one half-precision instruction has been added to this group.

Backports commit 70b4e6a445715519ae55179dc54f6e961ab30c27 from qemu
2018-03-08 23:43:52 -05:00
Alex Bennée b117df18df
arm/translate-a64: add FP16 FRSQRTE to simd_two_reg_misc_fp16
Backports commit c625ff95070e3ef96bd007de744e1d97c881efeb from qemu
2018-03-08 22:45:39 -05:00
Alex Bennée 068143595e
arm/helper.c: re-factor rsqrte and add rsqrte_f16
Much like recpe the ARM ARM has simplified the pseudo code for the
calculation which is done on a fixed point 9 bit integer maths. So
while adding f16 we can also clean this up to be a little less heavy
on the floating point and just return the fractional part and leave
the calle's to do the final packing of the result.

Backports commit d719cbc7641991d16b891ffbbfc3a16a04e37b9a from qemu

Also removes a load of symbols that seem unnecessary from the header_gen script
2018-03-08 22:42:04 -05:00
Alex Bennée fdb07713e6
arm/translate-a64: add FP16 FSQRT to simd_two_reg_misc_fp16
Backports commit b96a54c7e5576bd35b7d00d37b7929d2892d8cac from qemu
2018-03-08 21:57:35 -05:00
Alex Bennée 6102a61b14
arm/translate-a64: add FP16 FRCPX to simd_two_reg_misc_fp16
We go with the localised helper.

Backports commit 986950283837f697b35782b9ac3bc99fca614640 from qemu
2018-03-08 19:15:23 -05:00
Alex Bennée 4ea310c131
arm/translate-a64: add FP16 FRECPE
Now we have added f16 during the re-factoring we can simply call the
helper.

Backports commit fbd06e1e4b6566b4d727f9e553c819d034942f68 from qemu
2018-03-08 19:12:06 -05:00
Alex Bennée 5f3864c2c2
arm/helper.c: re-factor recpe and add recepe_f16
It looks like the ARM ARM has simplified the pseudo code for the
calculation which is done on a fixed point 9 bit integer maths. So
while adding f16 we can also clean this up to be a little less heavy
on the floating point and just return the fractional part and leave
the calle's to do the final packing of the result.

Backports commit 5eb70735af1c0b607bf2671a53aff3710cc1672f from qemu
2018-03-08 19:05:48 -05:00
Alex Bennée c590ff441c
arm/translate-a64: add FP16 FNEG/FABS to simd_two_reg_misc_fp16
Neither of these operations alter the floating point status registers
so we can do a pure bitwise operation, either squashing any sign
bit (ABS) or inverting it (NEG).

Backports commit 15f8a233c8c023dbc77b6fe6cd7c79eac9bee263 from qemu
2018-03-08 18:51:35 -05:00
Alex Bennée 7161c1ed52
arm/translate-a64: add FP16 SCVTF/UCVFT to simd_two_reg_misc_fp16 2018-03-08 18:48:25 -05:00
Alex Bennée 8ac9e3cff2
arm/translate-a64: add FP16 FCMxx (zero) to simd_two_reg_misc_fp16
I re-use the existing handle_2misc_fcmp_zero handler and tweak it
slightly to deal with the half-precision case.

Backports commit 7d4dd1a73a023f75c893623710e43743501b318e from qemu
2018-03-08 18:32:36 -05:00
Alex Bennée 39a68548d1
arm/translate-a64: add FCVTxx to simd_two_reg_misc_fp16
This covers all the floating point convert operations.

Backports commit 2df581304193d70eaf0d22cf4cb4613f74b6e59b from qemu
2018-03-08 18:25:29 -05:00
Alex Bennée d5f002b39a
arm/translate-a64: add FP16 FPRINTx to simd_two_reg_misc_fp16
This adds the full range of half-precision floating point to integral
instructions.

Backports commit 6109aea2d954891027acba64a13f1f1c7463cfac from qemu
2018-03-08 18:21:58 -05:00
Alex Bennée 33eda0f5d4
arm/translate-a64: initial decode for simd_two_reg_misc_fp16
This actually covers two different sections of the encoding table:

Advanced SIMD scalar two-register miscellaneous FP16
Advanced SIMD two-register miscellaneous (FP16)

The difference between the two is covered by a combination of Q (bit
30) and S (bit 28). Notably the FRINTx instructions are only
available in the vector form.

This is just the decode skeleton which will be filled out by later
patches.

Backports commit 5d432be6fd6efe37833ac82623c3abd35117b421 from qemu
2018-03-08 18:14:04 -05:00