Quantcast
Channel: Tools
Viewing all 91843 articles
Browse latest View live
↧

Forum Post: RE: CSS Is not automatically including the DSP/BIOS generated linker command file during linking

July 9, 2014, 1:57 pm
$
0
0

[quote user="Michael DiGiovanni"]

I have a project which had previously been building fine with C2000 compiler version 6.1.1, but we've recently upgraded to compiler version 6.2.6 and now it won't.

[/quote]

I'm not sure that just changing the compiler version alone triggered this issue. It is certainly possible but I couldn't reproduce it locally. I created a new BIOS5 project (same version as you are using) with CCS 5.2.1 using one of the examples. Built it to verify it built fine. Changed the compiler version from 6.1.0 to 6.2.6 and rebuilt, and there were not issues. The *cfg.cmd file was still getting passed to the linker correctly.

Did anything else change in your projetc other than the compiler version? Are you able to switch back to compiler 6.1.1 and make it build fine, and then switch back to 6.2.6 and make it fail?

↧
Search

Forum Post: RE: Port flash setting from flash.cmm to CCS5.x

July 9, 2014, 2:00 pm
$
0
0

Kirshna - since you also want to debug the application, I would stick solely with CCS.

I know very little (basically nothing) about Lauterbach *.cmm files so I can't really provide much help there but by taking a quick peek at it, it looks like it is configuring two regions for programming, both in internal flash. Internal flash is properly defined in the tms570ls3137.gel file. So starting a debug session for tms570ls3137 using that gel file (which should be used by default) will have set the debugger memory map correctly so you can flash your executable. I assume you are flashing an executable with debug symbols (*.out file) since you said you want to debug it and I can decipher from the *.cmm file that debug symbols and *.out files are referenced. Your executable should already have the information for the programmer to flash to the correct locations. While the debugger memory map will be properly configured by the gel file. 

Thanks

ki

↧
↧

Forum Post: RE: Cortex_M3_0: GEL: Target must be connected before calling the function

July 9, 2014, 2:14 pm
$
0
0

[quote user="Nahuel"]I either press the green bug button "debug" or I right click on the project -> debug as -> code composer studio[/quote]

The error was that the target was not connected before some GEL routine tried to access the target (I assume this was the program load request). The former action you do would try to auto-connect. Is the debugger actually connecting to the target correctly? When that error message occurs, is CCS connected to the target?

↧

Forum Post: RE: Type 'Int16' could not be resolved

July 9, 2014, 2:19 pm
↧

Forum Post: RE: No Chip Defined error #35

July 9, 2014, 2:38 pm
$
0
0

Yup.  That's the answer I needed.

Thanks.

↧
↧

Forum Post: RE: CCSv5 and CCSv6 occasionally hang on compiler error

July 9, 2014, 4:47 pm
$
0
0

Hi Ki-Soo

In our build server case, not specifying the -j and letting it figure out how many was bad news. I'm going to ask Somnath to try his PC with something like -j 4 instead of -j 8. Doing so seems to have helped another engineer in our group.

Wouldn't that be nice if it also helps the errors that never go away in the "Problems" tab? (wishful thinking perhaps)

Cheers
Eddie

↧

Forum Post: how to force double alignment to 4 bytes from gcc for ARM?

July 9, 2014, 5:03 pm
$
0
0

Hello,

I've run into a data alignment problem in my code running on the AM389x Sitara (ARM Cortex A8).  The code is compiled using the CodeSourcery gcc compiler.  

My problem is I have a structure that contains 13 int types followed by a double.  The double data in memory starts at the address immediately after the 13th int member.  I can see though that the code thinks the double starts at the address of the 13th int + 4 bytes (if I point to the double at the address the code tells me at less 4 bytes, it is printed correctly).  Alternately, if I add a dummy int member immediately before the double, the code prints the double correctly.

It looks like the processor aligns double on 4 byte boundaries (1 word), but the compiler thinks they are on 8 byte boundaries, and so assumes a 4 byte pad before the double after the 13 bytes.  As a solution, it seems the gcc option -mno-align-double would fix it, but the CodeSourcery gcc doesn't accept it (and it isn't listed in the CodeSourcery compiler manual under ARM options -- it is listed as an Intel 386 option).

Is there a way to make the compiler align doubles on 4 byte boundaries with ARM and this compiler?  Is there another way to fix this?  Inserting pads is not really a viable option because I have many structures like this and anybody could change one and ignore the need for the pad.

My data structure is shown below, followed by printfs of the start address of each structure member.  Also, __align_of__ (double) returns 8.

typedef struct input_timing_s
{
int chan;
int src_pwr_det;
int clk_det;
int de_det;
int sig_stable;
unsigned int time_locked;
int mode;
int hdcp_authenticated;
int hdcp_src_dev_auth_start;
int horiz_active;
int horiz_blank;
int vert_active;
int vert_blank;
double tmds_frequency;
int horiz_sync_polarity;
int vert_sync_polarity;
int interlace;
} input_timing_t;

addr of chan=0xbe9ca780
src_pwr_det=0xbe9ca784
clk_det=0xbe9ca788
de_det=0xbe9ca78c
sig_stable=0xbe9ca790
time_locked=0xbe9ca794
mode=0xbe9ca798
hdcp_authenticated=0xbe9ca79c
hdcp_src_dev_auth_start=0xbe9ca7a0
horiz_active=0xbe9ca7a4
horiz_blank=0xbe9ca7a8
vert_active=0xbe9ca7ac
vert_blank=0xbe9ca7b0
tmds_frequency=0xbe9ca7b8
horiz_sync_polarity=0xbe9ca7c0
vert_sync_polarity=0xbe9ca7c4
interlace=0xbe9ca7c8

Thanks,

Tim

↧

Forum Post: RE: Using Webnech Power Designer

July 9, 2014, 7:13 pm
$
0
0

Hi,

Sorry for the delay in reply, can you please mention your input voltage range?

Thanks, Sahil.

↧
Search

Forum Post: RE: CCS6 Looking up C Functions in Help

July 9, 2014, 9:27 pm
$
0
0

O.K. then...

How do I get this search Window to appear?

↧
↧

Forum Post: Installing CVS in CCSV6

July 9, 2014, 9:29 pm
$
0
0

I have been unable to get a CVS repository working on 3 different systems, which are running CCS V6.  One system, which had CCSV5 installed initially, and CCS V6 installed after, does have a CVS repository working.   Re-installing did not help. Installing CCSV5, then CCSV6 did not help.  I did not see any reference to this specific problem in CCS Forum.  I have been using CCS on these systems for about 6 months.

The system without CVS does NOT show "Eclipse CVS Client" under "Help>About > Installation Details",  The system with CVS working DOES show "ECLIPSE CVS Client" under "Help>About >Installation Details".

In CCS when selecting Window>Open Perspective>Other>Show All, CVS does NOT appear.  This is the common result on each system which does NOT have a CVS repository working.

I have NOT found a way to install a CVS plugin.  I did see a reference (somewhere) saying that CVS was a core part of Eclipse 4.3 and did not need to have a plugin installed.  I am not experienced with installing Eclipse plugins but that does not seem to be the real issue.

I have lots of existing projects including my present active project in CVS repository, so would prefer to not move to SVN or GIT at this time. I could revert to CCSV5 but do have a single system running CCSV6 with CVS working so reverting to V5 seems the wrong approach.

My "main development" computer running CCSV6 (without CVS working) is a Windows7 Professional OS, 64 bit, AMDFX-8350 8 core, 8gb ram.  The system with CCSV6 and CVS working is also WIn 7 Pro, 64 bit, AMD Phenom II X6.

Any advice will be appreciated,

Randy

↧

Forum Post: Adding external libray to CCSv5.5

July 9, 2014, 9:38 pm
$
0
0

Hi,

I am trying to add external library called armadillo (linear algebra library,www.arma.sourceforge.net) to CCSv5.5, I installed that library in my system (linux) and in CCS project->properties->C6000Compier->include options I gave path to folder under which that library got installed and added .so file and it's path in CCS project->properties->C6000linker->file search path. But when I build project I am getting errors in the header files (.hpp) of the library which I included. Is the way of adding this library properly done? If so why am I getting errors in header files of the library?.  BTW the library I am adding is C++ library.

↧

Forum Post: RE: Code size limitation

July 9, 2014, 10:11 pm
$
0
0

Hi AartiG,

 

Thank you very much for your reply.

After I correct the linker command file, there is only one code section left.

But there is another problem when I link a temporary .out file.

The .cmd file looks like:

-r
-x

d:\DEV\LCDIST\DISP\PTFSYST\BOOT\sbi_timer.obj
d:\DEV\LCDIST\DISP\PTFSYST\BOOT\sbs_atInit.obj
d:\DEV\LCDIST\DISP\PTFSYST\BOOT\sbs_atPerm.obj
d:\DEV\LCDIST\DISP\PTFSYST\MEMOIRE\smf2_drv.obj
d:\DEV\LCDIST\MAIN\OBJEMU_IEC61850\dm_ITMem.obj
d:\DEV\LCDIST\DISP\PTFSYST\MEMOIRE\smf_drv.obj
-l d:\DEV\LCDIST\MVP\UCA2\OBJEMU_IEC61850\mvp_IEC61850.lib
d:\DEV\LCDIST\DISP\PTFSYST\BOOT\sbi_tmi.obj
d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsuf_msf.obj
d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsu_Boot.obj
d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsu_Coproce.obj
d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsuf.obj
...
d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsuf_tr_cmde.obj

...
-l d:\DEV\LCDIST\MAIN\OBJEMU_IEC61850\main_IEC61850.lib

d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsuf_mvp.obj
d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsuf_mvp_pcf.obj
d:\DEV\LCDIST\SUPERVISION\OBJEMU_IEC61850\dsuf_mvp_str.obj

...
-l d:\DEV\LCDIST\DISP\LIV50300PL\util\util.lib
-l d:\DEV\LCDIST\lib\tex\rts.pc.lib
-l d:\DEV\LCDIST\lib\tex\mathrco2acfim.lib


SECTIONS
{
    .text:
    {
        *(.text)
    }
}

Then I adjust the location of the overflowed .obj file. If I am lucky, the overflow problem may disappear, if not, another overflow problem will appear.

The next time, when we fix some bugs in the code, we link all libraries and .obj files again, then we will encounter some other overflow problems.

How can we avoid this from happening? Is there any way to let the overflow problem never show again?

Thanks!

↧

Forum Post: RE: GNU compiler in CCSv6

July 9, 2014, 10:29 pm
$
0
0

I installed TI-RTOS vor TivaC. Now I get the error : device not supported?

The compiler then lists all M4 devices? Why not the M3?

 

↧
↧

Forum Post: Multiple MSP430 targets to single GUI

July 9, 2014, 10:48 pm
$
0
0

Hi,

I am working with 3 DRV8711-Booster packs along with MSP430G2553 launchpads to run 3 stepper motors. I made a GUI using GUI Composer. I made three different .out files with different variables for each kit. Then I exported the GUI (with variables of all three kits). All launchpads are on different COM ports and I made three separate configuration .ccxml files (with unique COM Port hardcoded). In the appInit.js script, I first set config for launchpad 1 (using its .ccxml file) and loads its symbols (through its .out file), then set config for launchpad 2 and loads its symbol and similarly for launchpad 3. GUI loads, but I can only control 1 motor (the last one).

I am using CCSv6. I want to control each motor together. Help required.

Regards

Aman

↧

Forum Post: printf + heap : difference IAR-CCS

July 9, 2014, 11:35 pm
$
0
0

Hello,

When the heap size =0, does anybody know why the printf fonction works correctly with IAR while it does not work with CCS?


Best regards

Mich

↧
Search

Forum Post: RE: Debugging SYS BIOS project

July 10, 2014, 12:40 am
$
0
0

According to one of my colleagues, it might be because my kernel isn't loading the task? If that's the case how do I actually set it such that my kernel actually runs it? How do I access the kernel?

↧

Forum Post: RE: GNU compiler in CCSv6

July 10, 2014, 12:35 pm
$
0
0

OK, I will. Thanks for the effort.

↧
↧

Forum Post: RE: TMS570 (TMS570LS20216): Linker generated ECC issue

July 10, 2014, 1:40 pm
$
0
0

Thanks for looking into this. It turns out that it was working all along and that the compiler optimization (-O2) was preventing execution of the interrupt handler (I didn't look for the root cause of this).  Turning off optimizations on the build fixed the issue. Regardless, it was helpful to get verification of the linker script.  Thanks again!

↧

Forum Post: RE: CSS Is not automatically including the DSP/BIOS generated linker command file during linking

July 10, 2014, 2:18 pm
$
0
0

I'm not sure what more we can do without the build log or the project. It is ofcourse possible that it is a bug specific to CCS 5.2.1 and your project (especially since your co-workers don't see the issue with 5.2.0) but I'm afraid it is difficult to determine the reason why the BIOS generated command file is not being pulled into the link without further details.

↧

Forum Post: RE: DMA setting problem.

July 10, 2014, 2:27 pm
$
0
0

There are a couple of things going on here.

Main thing first is that it is invalid to cast a data pointer to a function pointer or vice versa (even an explicit cast).  Hence the warning about “invalid type conversion”.

For DMA setup on an mspx device, we need to get a 20-bit address.  This requires using the intrinsic:

                __data16_write_addr(unsigned short addr, unsigned long src)

The intrinsic generates the desired sequence including MOVA instruction:

                MOV.W addr, Rx

                MOVA src, 0(Rx)

 Please see msp430 dma example code for details: http://www.ti.com/lsds/ti/microcontrollers_16-bit_32-bit/msp/tools_software.page#code

Or if you have MSP430Ware installed, you can browse through MSP430Ware content inside TI Resource Explorer to find the code examples.

Regards,
Greg

↧
Viewing all 91843 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>