|
We use 21161 for our project. and VisualDSP++ 2.0 We encountered a strange situation in one of our code files When we compile the code with the -Os switch, the compiler produces a code which tries to utilise the SIMD(Single Instruction Multiple data) feature in the 21161 processor. This code works in some cases and does not work in some other cases. The following example does not work properly. static void xyz ( float *InReal, /* in, Real part of FFT */ float *InImag, /* in, Imaginary part of FFT */ float *Output ) /* out, Spectre de puissance */ { int i; /* needs to be volatile to stop optimizer using the SIMD feture or -no-simd compiler option has to be used */ /* Calcul de la raie 0 */ Output[0] = InReal[0]*InReal[0]; /* Calcul du spectre de puissance */ for (i=1; i<8; i++) { Output[i] = InReal[i]*InReal[i] + InImag[i]*InImag[i]; } } and a call to this function with sample data does not prodece expected data. float InReal[8] = { 0,1.11,2.22,3.33,4.44,5.55,6.66,7.77 }; float InImag[8] = { 0,1.11,2.22,3.33,4.44,5.55,6.66,7.77 }; float Output[8]; xyz(InReal,InImag,Output); In the above example, the code in the "for" loop was optimised to utilise the SIMD feature of the processor. The assembler code looks like bit set mode1 0x200000; nop; m4=2; -- bubble -- r2=dm(i4,m4); F2F2, r12=dm(i0,m4); lcntr=3, do(pc,.P2L4-1)until lce; . F12*F12, r8=dm(i4,m4); F2F12, r12=dm(i0,m4); F2F8, dm(i1,2)=r2; //end loop .P2L3; .P2L4: bit clr mode1 0x200000; nop; Though the PEYEN bit is set in the MODE1 Register, the code does not seem to work as expected. The computations inside the "for" loop are supposed to be performed in parallel in the two processing elements (PEx and PEy) with different sets of data. This situation was simulated in the VDSP simulator and it produces the expected result. When it was tried in the emulator, only onset of computations (the explicit ones on the PEx) seem to be performed and the computations on the implicit registers, PEy, seems to contain zeros. And hence only the alternate elements in the "Output" array get meaningful data and the remaining entries contain zeros. Any suggestions? Is there any thing else we need to take care to enable SIMD mode apart from the MODE1 register? Thanks, Kannan, Contract Engineer, TRW Automotive Electronics, Cirencester, UK Gl7 1EN |
|
|
21161/VDSP 2.0++ SIMD mode
Started by ●October 31, 2002
Reply by ●November 1, 20022002-11-01
|
--On Thursday, October 31, 2002 6:45 PM +0000 Kannan Subramanian-nonTRW <> wrote: > We use 21161 for our project. and VisualDSP++ 2.0 But which *release* of VDSP? The latest release was 6.1.9. ("cc21k -version" will tell you the current release.) Lots of bugs have been fixed since the original release (6.1.0) and I'm pretty sure some had to do with SIMD. Releases are available as zip files on ADI's FTP site. There's no nice installer, so you need to unzip them to your VisualDSP directory, maintaining the directory structure. I recommend backup up the old directory first in case you need to revert to the prior release. Bugs *do* sometimes get introduced in interim releases, so they're not widely advertised until users who really need them have had a chance to test them. (Given a reproducable bug, ADI can get a new release out pretty quick, and will normally have a coding workaround in the meantime.) |
Reply by ●November 1, 20022002-11-01
|
Thanks for the suggestions. I tried with the new patch from the ADI ftp site i.e., release 6.1.9.1 but it does not solve the problem. thanks, Kannan Contract Engineer, TRW Automotive, Cirencester UK >>> Kenneth Porter <> 11/01/02 12:55am >>> --On Thursday, October 31, 2002 6:45 PM +0000 Kannan Subramanian-nonTRW <> wrote: > We use 21161 for our project. and VisualDSP++ 2.0 But which *release* of VDSP? The latest release was 6.1.9. ("cc21k -version" will tell you the current release.) Lots of bugs have been fixed since the original release (6.1.0) and I'm pretty sure some had to do with SIMD. Releases are available as zip files on ADI's FTP site. There's no nice installer, so you need to unzip them to your VisualDSP directory, maintaining the directory structure. I recommend backup up the old directory first in case you need to revert to the prior release. Bugs *do* sometimes get introduced in interim releases, so they're not widely advertised until users who really need them have had a chance to test them. (Given a reproducable bug, ADI can get a new release out pretty quick, and will normally have a coding workaround in the meantime.) _____________________________________ Note: If you do a simple "reply" with your email client, only the author of this message will receive your answer. You need to do a "reply all" if you want your answer to be distributed to the entire group. _____________________________________ About this discussion group: To Join: Send an email to To Post: Send an email to To Leave: Send an email to Archives: http://groups.yahoo.com/group/adsp Other Groups: http://www.dsprelated.com/groups.php3 ">http://docs.yahoo.com/info/terms/ |
|
|
Reply by ●November 2, 20022002-11-02
|
--On Friday, November 01, 2002 09:48:44 AM +0000 Kannan
Subramanian-nonTRW <> wrote: > I tried with the new patch from the ADI ftp site i.e., release 6.1.9.1 > but it does not solve the problem. The next step is to put together a test case (source file, command line, steps to reproduce) and submit it to the support address. (The easier you make it for them to *see* the problem on *their* setup, the faster they fix the bug.) Feel free to cc me on the support request. Their system is smart enough to know to reply to all. |
