two question

1-When I read "3.1.5 CPX and PCX Packet Formats" in OpenSPARCT1_Micro_Arch.pdf,The document says the "ERR" bits are 138:137, and 139 bit is L2 miss

But I found CPX description in iop.h file like this:

`define CPX_ERR_HI 139

`define CPX_ERR_LO 137

It's conflict with the doc's description.

Who can tell me,which is correct?

2- When I use regreport script to check result of core1 full regression,I found that one test failed.It's puzzled me.Have any one has run this regression?Your result is same with mine,or I have do some error in the sim flow?

[593 byte] By [sunboy_cja] at [2007-11-27 8:57:57]

«« Where can i get swing source code?
»» warning: [path] Unexpected file on path:

# 1

There is actually no conflict between the OpenSPARCT1_Micro_Arch.pdf and the iop.h include file. It's just a difference in terminology. The iop.h include file defines bits 139:137 of the CPX packet as an error field. Bit 139 indicates L2 miss, bit 138 indicates an un-correctible ECC error in L2, and bit 137 indicates a correctible ECC error in L2.

Since an L2 miss is not technically an error, the Micro-Arch spec separates out the L2 Miss bit (bit 139) as a separate field, and calls bit 138:137 the error field.

As to your failing test: Which test is failing? Perhaps I can take a look.

Hope this helps

formalGuya at 2007-7-12 21:23:13 >

top of Java-index,Open Source Technologies,OpenSPARC...

# 2

Thank you for your reply!I have not modified the rtl source code when Runing regression with VCS.My failed test name is tlu_stb2b_trap_60.sreason:Hit bad trap.name: bug3138.sreason : timeout

sunboy_cja at 2007-7-12 21:23:13 >

OpenSPARC

Java programming resources: FAQs, tutorials, compiler and browser download sites, documentation, books lists, IDEs, etc.

two question

Open Source Technologies Hot Topic

Open Source Technologies

All Classified