SWCU195A December 2024 – May 2025 CC2744R7-Q1 , CC2745P10-Q1 , CC2745R10-Q1 , CC2745R7-Q1 , CC2755P10 , CC2755R10
Table 2-135 lists the memory-mapped registers for the FPE registers. All register offset addresses not listed in Table 2-135 should be considered as reserved locations and the register contents should not be modified.
| Offset | Acronym | Register Name | Section |
|---|---|---|---|
| 4h | FPCCR | Holds control data for the Floating-point extension | Section 2.7.7.1 |
| 8h | FPCAR | Holds the location of the unpopulated floating-point register space allocated on an exception stack frame | Section 2.7.7.2 |
| Ch | FPDSCR | Holds the default values for the floating-point status control data that the PE assigns to the FPSCR when it creates a new floating-point context | Section 2.7.7.3 |
| 10h | MVFR0 | Describes the features provided by the Floating-point Extension | Section 2.7.7.4 |
| 14h | MVFR1 | Describes the features provided by the Floating-point Extension | Section 2.7.7.5 |
| 18h | MVFR2 | Describes the features provided by the Floating-point Extension | Section 2.7.7.6 |
Complex bit access types are encoded to fit into small table cells. Table 2-136 shows the codes that are used for access types in this section.
| Access Type | Code | Description |
|---|---|---|
| Read Type | ||
| R | R | Read |
| Write Type | ||
| W | W | Write |
| Reset or Default Value | ||
| -n | Value after reset or the default value | |
FPCCR is shown in Table 2-137.
Return to the Summary Table.
Holds control data for the Floating-point extension
| Bit | Field | Type | Reset | Description |
|---|---|---|---|---|
| 31 | ASPEN | R/W | 1h | When this bit is set to 1, execution of a floating-point instruction sets the CONTROL.FPCA bit to 1 |
| 30 | LSPEN | R/W | 1h | Enables lazy context save of floating-point state |
| 29 | LSPENS | R/W | 0h | This bit controls whether the LSPEN bit is writeable from the Non-secure state |
| 28 | CLRONRET | R/W | 0h | Clear floating-point caller saved registers on exception return |
| 27 | CLRONRETS | R/W | 0h | This bit controls whether the CLRONRET bit is writeable from the Non-secure state |
| 26 | TS | R/W | 0h | Treat floating-point registers as Secure enable |
| 25-11 | RES0 | R | 0h | Reserved, RES0 |
| 10 | UFRDY | R/W | 0h | Indicates whether the software executing when the PE allocated the floating-point stack frame was able to set the UsageFault exception to pending |
| 9 | SPLIMVIOL | R/W | 0h | This bit is banked between the Security states and indicates whether the floating-point context violates the stack pointer limit that was active when lazy state preservation was activated. SPLIMVIOL modifies the lazy floating-point state preservation behavior |
| 8 | MONRDY | R/W | 0h | Indicates whether the software executing when the PE allocated the floating-point stack frame was able to set the DebugMonitor exception to pending |
| 7 | SFRDY | R/W | 0h | Indicates whether the software executing when the PE allocated the floating-point stack frame was able to set the SecureFault exception to pending. This bit is only present in the Secure version of the register, and behaves as RAZ/WI when accessed from the Non-secure state |
| 6 | BFRDY | R/W | 0h | Indicates whether the software executing when the PE allocated the floating-point stack frame was able to set the BusFault exception to pending |
| 5 | MMRDY | R/W | 0h | Indicates whether the software executing when the PE allocated the floating-point stack frame was able to set the MemManage exception to pending |
| 4 | HFRDY | R/W | 0h | Indicates whether the software executing when the PE allocated the floating-point stack frame was able to set the HardFault exception to pending |
| 3 | THREAD | R/W | 0h | Indicates the PE mode when it allocated the floating-point stack frame |
| 2 | S | R/W | 0h | Security status of the floating-point context. This bit is only present in the Secure version of the register, and behaves as RAZ/WI when accessed from the Non-secure state. This bit is updated whenever lazy state preservation is activated, or when a floating-point instruction is executed |
| 1 | USER | R/W | 0h | Indicates the privilege level of the software executing when the PE allocated the floating-point stack frame |
| 0 | LSPACT | R/W | 0h | Indicates whether lazy preservation of the floating-point state is active |
FPCAR is shown in Table 2-138.
Return to the Summary Table.
Holds the location of the unpopulated floating-point register space allocated on an exception stack frame
| Bit | Field | Type | Reset | Description |
|---|---|---|---|---|
| 31-3 | ADDRESS | R/W | 0h | The location of the unpopulated floating-point register space allocated on an exception stack frame |
| 2-0 | RES0 | R | 0h | Reserved, RES0 |
FPDSCR is shown in Table 2-139.
Return to the Summary Table.
Holds the default values for the floating-point status control data that the PE assigns to the FPSCR when it creates a new floating-point context
| Bit | Field | Type | Reset | Description |
|---|---|---|---|---|
| 31-27 | RES0 | R | 0h | Reserved, RES0 |
| 26 | AHP | R/W | 0h | Default value for FPSCR.AHP |
| 25 | DN | R/W | 0h | Default value for FPSCR.DN |
| 24 | FZ | R/W | 0h | Default value for FPSCR.FZ |
| 23-22 | RMode | R/W | 0h | Default value for FPSCR.RMode |
| 21-0 | RES0_1 | R | Xh | Reserved, RES0 |
MVFR0 is shown in Table 2-140.
Return to the Summary Table.
Describes the features provided by the Floating-point Extension
| Bit | Field | Type | Reset | Description |
|---|---|---|---|---|
| 31-28 | FPRound | R | 1h | Indicates the rounding modes supported by the FP Extension |
| 27-24 | RES0 | R | 0h | Reserved, RES0 |
| 23-20 | FPSqrt | R | 1h | Indicates the support for FP square root operations |
| 19-16 | FPDivide | R | 1h | Indicates the support for FP divide operations |
| 15-12 | RES0_1 | R | 0h | Reserved, RES0 |
| 11-8 | FPDP | R | 0h | Indicates support for FP double-precision operations |
| 7-4 | FPSP | R | 2h | Indicates support for FP single-precision operations |
| 3-0 | SIMDReg | R | 1h | Indicates size of FP register file |
MVFR1 is shown in Table 2-141.
Return to the Summary Table.
Describes the features provided by the Floating-point Extension
| Bit | Field | Type | Reset | Description |
|---|---|---|---|---|
| 31-28 | FMAC | R | 1h | Indicates whether the FP Extension implements the fused multiply accumulate instructions |
| 27-24 | FPHP | R | 1h | Indicates whether the FP Extension implements half-precision FP conversion instructions |
| 23-8 | RES0 | R | 0h | Reserved, RES0 |
| 7-4 | FPDNaN | R | 1h | Indicates whether the FP hardware implementation supports NaN propagation |
| 3-0 | FPFtZ | R | 1h | Indicates whether subnormals are always flushed-to-zero |
MVFR2 is shown in Table 2-142.
Return to the Summary Table.
Describes the features provided by the Floating-point Extension
| Bit | Field | Type | Reset | Description |
|---|---|---|---|---|
| 31-8 | RES0 | R | Xh | Reserved, RES0 |
| 7-4 | FPMisc | R | 4h | Indicates support for miscellaneous FP features |
| 3-0 | RES0_1 | R | 0h | Reserved, RES0 |