Atmel Xplained Pro Evaluation Kit ATSAM4E-XPRO ATSAM4E-XPRO Fiche De Données

Codes de produits
ATSAM4E-XPRO
Page de 1506
397
SAM4E [DATASHEET]
Atmel-11157D-ATARM-SAM4E16-SAM4E8-Datasheet_12-Jun-14
21.4.2 Read Operations
An optimized controller manages embedded Flash reads, thus increasing performance when the processor is
running in Thumb-2 mode by means of the 128- or 64-bit-wide memory interface.
The Flash memory is accessible through 8-, 16- and 32-bit reads.
As the Flash block size is smaller than the address space reserved for the internal memory area, the embedded
Flash wraps around the address space and appears to be repeated within it.
The read operations can be performed with or without wait states. Wait states must be programmed in the field
FWS (Flash Read Wait State) in the Flash Mode register (EEFC_FMR). Defining FWS as 0 enables the single-
cycle access of the embedded Flash. Refer to the Electrical Characteristics section for more details.
21.4.2.1 128-bit or 64-bit Access Mode
By default, the read accesses of the Flash are performed through a 128-bit wide memory interface. It improves
system performance especially when two or three wait states are needed.
For systems requiring only 1 wait state, or to focus on current consumption rather than performance, the user can
select a 64-bit wide memory access via the FAM bit in EEFC_FMR.
Refer to the Electrical Characteristics section for more details.
21.4.2.2 Code Read Optimization
Code read optimization is enabled if the SCOD bit in EEFC_FMR is cleared.
A system of 2 x 128-bit or 2 x 64-bit buffers is added in order to optimize sequential code fetch. 
Note:
Immediate consecutive code read accesses are not mandatory to benefit from this optimization.
The sequential code read optimization is enabled by default. If the SCOD bit in EEFC_FMR is set to 1, these
buffers are disabled and the sequential code read is no longer optimized.
Another system of 2 x 128-bit or 2 x 64-bit buffers is added in order to optimize loop code fetch. Refer to “Code
Loop Optimization” on page 398 for
 more details. 
Figure 21-2.
Code Read Optimization for FWS = 0 
Note: When FWS is equal to 0, all the accesses are performed in a single-cycle access.
Flash Access
Buffer 0 (128bits)
Master Clock
ARM Request
       (32-bit)
XXX
Data To ARM
Bytes 0-15
Bytes 16-31
Bytes 32-47
Bytes 0-15
Buffer 1 (128bits)
Bytes 32-47
Bytes 0-3
Bytes 4-7
Bytes 8-11
Bytes 12-15
Bytes 16-19
Bytes 20-23
Bytes 24-27
XXX
XXX
Bytes 16-31
@Byte 0
@Byte 4
@Byte 8
@Byte 12
@Byte 16
@Byte 20
@Byte 24
@Byte 28
@Byte 32
Bytes 28-31