* [Qemu-devel] [PATCH v2] target-arm: implement CPACR register logic for ARMv7
@ 2014-05-19 20:56 Fabian Aggeler
2014-05-23 13:14 ` Peter Maydell
0 siblings, 1 reply; 2+ messages in thread
From: Fabian Aggeler @ 2014-05-19 20:56 UTC (permalink / raw)
To: qemu-devel; +Cc: peter.maydell, peter.crosthwaite
In ARMv7 the CPACR register allows to control access rights to
coprocessor 0-13 interfaces. Bits corresponding to unimplemented
coprocessors should be RAZ/WI. Bits ASEDIS, D32DIS, TRCDIS are
UNK/SBZP if VFP is not implemented and RAO/WI in some cases.
Treating TRCDIS as RAZ/WI since we neither implement a trace
macrocell nor a CP14 interface to the trace macrocell registers.
Since CPACR bits for VFP/Neon access are honoured with the CPACR_FPEN
bit in the TB flags, flushing the TLB is not necessary anymore.
Signed-off-by: Fabian Aggeler <aggelerf@ethz.ch>
---
v1 -> v2:
* Removed unnecessary checks in disas_coproc_insn()
* Made changes in cpacr_write() ARMv7 specific
* Accounting for ASEDIS, D32DIS and TRCDIS bits
target-arm/helper.c | 32 ++++++++++++++++++++++++++++----
1 file changed, 28 insertions(+), 4 deletions(-)
diff --git a/target-arm/helper.c b/target-arm/helper.c
index 417161e..cb59f00 100644
--- a/target-arm/helper.c
+++ b/target-arm/helper.c
@@ -477,11 +477,35 @@ static const ARMCPRegInfo not_v7_cp_reginfo[] = {
static void cpacr_write(CPUARMState *env, const ARMCPRegInfo *ri,
uint64_t value)
{
- if (env->cp15.c1_coproc != value) {
- env->cp15.c1_coproc = value;
- /* ??? Is this safe when called from within a TB? */
- tb_flush(env);
+ uint32_t mask = 0;
+
+ /* In ARMv8 most bits of CPACR_EL1 are RES0. */
+ if (!arm_feature(env, ARM_FEATURE_V8)) {
+ /* ARMv7 defines bits for unimplemented coprocessors as RAZ/WI.
+ * ASEDIS [31] and D32DIS [30] are both UNK/SBZP without VFP.
+ * TRCDIS [28] is RAZ/WI since we do not implement a trace macrocell.
+ */
+ if (arm_feature(env, ARM_FEATURE_VFP)) {
+ /* VFP coprocessor: cp10 & cp11 [23:20] */
+ mask |= (1 << 31) | (1 << 30) | (0xf << 20);
+
+ if (!arm_feature(env, ARM_FEATURE_NEON)) {
+ /* ASEDIS [31] bit is RAO/WI */
+ value |= (1 << 31);
+ }
+
+ /* VFPv3 and upwards with NEON implement 32 double precision
+ * registers (D0-D31).
+ */
+ if (!arm_feature(env, ARM_FEATURE_NEON) ||
+ !arm_feature(env, ARM_FEATURE_VFP3)) {
+ /* D32DIS [30] is RAO/WI if D16-31 are not implemented. */
+ value |= (1 << 30);
+ }
+ }
+ value &= mask;
}
+ env->cp15.c1_coproc = value;
}
static const ARMCPRegInfo v6_cp_reginfo[] = {
--
1.8.3.2
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [Qemu-devel] [PATCH v2] target-arm: implement CPACR register logic for ARMv7
2014-05-19 20:56 [Qemu-devel] [PATCH v2] target-arm: implement CPACR register logic for ARMv7 Fabian Aggeler
@ 2014-05-23 13:14 ` Peter Maydell
0 siblings, 0 replies; 2+ messages in thread
From: Peter Maydell @ 2014-05-23 13:14 UTC (permalink / raw)
To: Fabian Aggeler; +Cc: Peter Crosthwaite, QEMU Developers
On 19 May 2014 21:56, Fabian Aggeler <aggelerf@ethz.ch> wrote:
> In ARMv7 the CPACR register allows to control access rights to
> coprocessor 0-13 interfaces. Bits corresponding to unimplemented
> coprocessors should be RAZ/WI. Bits ASEDIS, D32DIS, TRCDIS are
> UNK/SBZP if VFP is not implemented and RAO/WI in some cases.
> Treating TRCDIS as RAZ/WI since we neither implement a trace
> macrocell nor a CP14 interface to the trace macrocell registers.
>
> Since CPACR bits for VFP/Neon access are honoured with the CPACR_FPEN
> bit in the TB flags, flushing the TLB is not necessary anymore.
>
> Signed-off-by: Fabian Aggeler <aggelerf@ethz.ch>
> ---
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
I'll add this into the target-arm.next tree.
thanks
-- PMM
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2014-05-23 13:14 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-05-19 20:56 [Qemu-devel] [PATCH v2] target-arm: implement CPACR register logic for ARMv7 Fabian Aggeler
2014-05-23 13:14 ` Peter Maydell
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).