boot.asm source for Tiva 1294 build?

If the application is built for TI-RTOS then I would expect the disassembly view and source correlation to automatically happen with the boot source file within TI-RTOS. Can you show a screenshot of what you see in the disassembly view?

What does the link map file show - is the boot code coming from TI-RTOS or the compiler runtime library? Has the .out file or symbols for the .out file been loaded on to the target?

Here is a screen shot. You can see that the disassembler thinks it is at line 171 of boot.asm with an instruction "LDR r0, c_stack" when the actual machine code is loading 0xed88 into R1.

Here is the relevant section from the map file:

                  00013c18    00000058     stk.aem4f : bind.oem4f (.text:BindFree)
                  00013c70    00000058     boot.aem4f : boot.oem4f (.text)
                  00013cc8    00000058     driverlib.lib : emac.obj (.text:EMACPHYConfigSet)
                  00013d20    00000058     stk.aem4f : igmp.oem4f (.text:IGMPTimer)

You can see that the boot.asm code is somehow mangled into a file called boot.aem4f that *never* exists where a real person can actually go see what was really built into the system. This behavior is incredibly annoying when things go wrong. Unfortunately, "things going wrong" is not an infrequent occurrence.

Can TI please, please, please change the XDC tool behavior to save *ALL* source code that is generated as well as *ALL* build scripts so we can go see what the system actually did when things don't make sense? There are many thousands of steps and decisions that are made during the build process and we as users have no way of figuring out what got screwed up.

In the case of the file under consideration, we should see the actual file as mangled by the XDC tools and have a listing of the generated code.

Yes the symbols were loaded. You can see that because CCSv6 actually knows that we are at _c_int00. It just thinks the source code is somewhere else.

It also looks very much like none of the boot.asm files in my tree are what is actually used to build my image.  There are two versions in the ccsv6 tools path and two in the tirtos path and it appears that none of them is actually what is in my image.

The closest version of boot.asm that seems to have been used appears to be the one from ti\tirtos_tivac_2_00_01_23\products\bios_6_40_01_15\packages\ti\targets\arm\rtsarm.  The code that exits the assembler preprocessor actually appears to start at line 159 (conditional assembly becomes TRUE at the .else) with actual code beginning at line 169. However, you can see from the disassembly that CCSv6 incorrectly thinks it is at line 171.

   .else   (Line 159)
    .if __TI_TMS470_V7M4__ & __TI_VFP_SUPPORT__

    .thumb
    ;*------------------------------------------------------
    ;* SETUP FULL ACCESS TO COPROCESSORS 10 AND 11,
    ;* REQUIRED FOR FP. COPROCESSOR ACCESS CONTROL REG
    ;* BITS [23:22] - CP11, [21:20] - CP10
    ;* SET TO 0b11 TO ENABLE FULL ACCESS
        ;*------------------------------------------------------
cpacr   .set     0xE000ED88            ; CAPCR address
    MOVW     r1, #cpacr & 0xFFFF (Line 169)
    MOVT     r1, #cpacr >> 16
    LDR      r0, [ r1 ]

OK, the screen shot should come through now.

Rats. Wrong file.  ANother try.