| Electronic Components Datasheet Search |
|
MPC8540 Datasheet(PDF) 339 Page - Freescale Semiconductor, Inc |
|
|
|||||||||||||||||||||||||||||
MPC8540 Datasheet(HTML) 339 Page - Freescale Semiconductor, Inc |
|
339 / 1302 page ![]() Initialization/Application Information MPC8540 PowerQUICC III Integrated Host Processor Reference Manual, Rev. 1 Freescale Semiconductor 7-35 7.10.1 Initialization This section describes L1 cache and memory-mapped SRAM initialization. 7.10.1.1 L2 Cache Initialization After power-on reset, the valid bits in the L2 cache status array are in random states. Therefore, it is necessary to perform a flash invalidate operation before using the array as an L2 cache. This is done by writing a 1 to the L2I bit of the L2 control register (L2CTL). This can be done before or simultaneously with the write that enables the L2 cache. That is, the L2E and L2I bits of L2CTL can be set simultaneously. The L2I bit clears automatically, so no further writes are necessary. 7.10.1.2 Memory-Mapped SRAM Initialization After power-on reset, the contents of the data and ECC arrays and are random, so all SRAM data must be initialized before it is read. If the cache is initialized by the core or any other device that uses sub-cacheline transactions, ECC error checking should be disabled during the initialization process to avoid false ECC errors generated during the read-modify-write process used for sub-cacheline writes to the SRAM array. This is done by setting the multi- and single-bit ECC error disable bits of the L2 error disable register (L2ERRDIS[MBECCDIS, SBECCDIS]). See Section 7.3.1.5.2, “Error Control and Capture Registers”. If the array is initialized by a DMA engine using cache-line writes, then ECC checking can remain enabled during the initialization process. 7.10.2 Managing Errors This section describes recommended handling for ECC and tag parity errors. 7.10.2.1 ECC Errors An individual soft error that causes a single- or mulit-bit ECC error can be cleared from the L2 array simply by executing a dcbf instruction for the address captured in the L2ERRADDR register. This will invalidate the line in the L2 cache. When the load that caused the ECC error is performed again, the data will be re-allocated into the L2 with ECC bits set properly again. If the threshold for single bit errors set in the L2ERRCTL register is exceeded, then the L2 cache should be flash invalidated to clear out all single-bit errors. Note that no data is lost by executing dcbf instructions or flash invalidate operations because the L2 cache is write-through and contains no modified data. |
|
Link URL |
| Does ALLDATASHEET help your business so far? [ DONATE ] |
About Alldatasheet | Advertisement | Contact us | Privacy Policy | Link to Datasheet | Link Exchange | Manufacturer List All Rights Reserved©Alldatasheet.com |
| Russian : Alldatasheetru.com | Korean : Alldatasheet.co.kr | Spanish : Alldatasheet.es | French : Alldatasheet.fr | Italian : Alldatasheetit.com Portuguese : Alldatasheetpt.com | Polish : Alldatasheet.pl | Vietnamese : Alldatasheet.vn Indian : Alldatasheet.in | Mexican : Alldatasheet.com.mx | British : Alldatasheet.co.uk | New Zealand : Alldatasheet.co.nz |
|
Family Site : ic2ic.com |
icmetro.com |