Recherche avancée

Sur d’autres sites (6132)

• Bit-field badness

  30 janvier 2010, par Mans — Compilers, Optimisation

  Consider the following C code which is based on an real-world situation.

  struct bf1_31 
      unsigned a:1 ;
      unsigned b:31 ;
   ;
  void func(struct bf1_31 *p, int n, int a)
  
  
  
      int i = 0 ;
  
      do 
  
          if (p[i].a)
  
              p[i].b += a ;
  
       while (++i < n) ;
  
  
  

  How would we best write this in ARM assembler ? This is how I would do it :
  

  func :
          ldr     r3,  [r0], #4
          tst     r3,  #1
          add     r3,  r3,  r2,  lsl #1
          strne   r3,  [r0, #-4]
          subs    r1,  r1,  #1
          bgt     func
          bx      lr
  

  The add instruction is unconditional to avoid a dependency on the comparison. Unrolling the loop would mask the latency of the ldr instruction as well, but that is outside the scope of this experiment.

  Now compile this code with gcc -march=armv5te -O3 and watch in horror :

  func :
          push    r4
          mov     ip, #0
          mov     r4, r2
  loop :
          ldrb    r3, [r0]
          add     ip, ip, #1
          tst     r3, #1
          ldrne   r3, [r0]
          andne   r2, r3, #1
          addne   r3, r4, r3, lsr #1
          orrne   r2, r2, r3, lsl #1
          strne   r2, [r0]
          cmp     ip, r1
          add     r0, r0, #4
          blt     loop
          pop     r4
          bx      lr
  

  This is nothing short of awful :

  • The same value is loaded from memory twice.
  • A complicated mask/shift/or operation is used where a simple shifted add would suffice.
  • Write-back addressing is not used.
  • The loop control counts up and compares instead of counting down.
  • Useless mov in the prologue ; swapping the roles or r2 and r4 would avoid this.
  • Using lr in place of r4 would allow the return to be done with pop {pc}, saving one instruction (ignoring for the moment that no callee-saved registers are needed at all).

  Even for this trivial function the gcc-generated code is more than twice the optimal size and slower by approximately the same factor.

  The main issue I wanted to illustrate is the poor handling of bit-fields by gcc. When accessing bitfields from memory, gcc issues a separate load for each field even when they are contained in the same aligned memory word. Although each load after the first will most likely hit L1 cache, this is still bad for several reasons :

  • Loads have typically two or three cycles result latency compared to one cycle for data processing instructions. Any bit-field can be extracted from a register with two shifts, and on ARM the second of these can generally be achieved using a shifted second operand to a following instruction. The ARMv6T2 instruction set also adds the SBFX and UBFX instructions for extracting any signed or unsigned bit-field in one cycle.
  • Most CPUs have more data processing units than load/store units. It is thus more likely for an ALU instruction than a load/store to issue without delay on a superscalar processor.
  • Redundant memory accesses can trigger early flushing of store buffers rendering these less efficient.

  No gcc bashing is complete without a comparison with another compiler, so without further ado, here is the ARM RVCT output (armcc --cpu 5te -O3) :

  func :
          mov     r3, #0
          push    r4, lr
  loop :
          ldr     ip, [r0, r3, lsl #2]
          tst     ip, #1
          addne   ip, ip, r2, lsl #1
          strne   ip, [r0, r3, lsl #2]
          add     r3, r3, #1
          cmp     r3, r1
          blt     loop
          pop     r4, pc
  

  This is much better, the core loop using only one instruction more than my version. The loop control is counting up, but at least this register is reused as offset for the memory accesses. More remarkable is the push/pop of two registers that are never used. I had not expected to see this from RVCT.

  Even the best compilers are still no match for a human.

• Real time stereo video stream. How to start

  29 juin 2015, par victor jung

  Here is my problem :

  I have at one side two webcams plugged in a linux computer, and on the other side, an android smartphone. My goal is to program a real time stream of the 2 webcams, and display that stream on the smartphone (to be used in a google cardboard mask). I read quite a lot on the subject, and I found several way to achieve this.

  First there is ffmpeg, which could encode and stream.

  But I need to modify the pictures at some point, to re-size them / or to ad some kind of distortion. So I thought it would be great if I can get the 2 images, with openCv maybe, playwith them, build a new one, and stream them, but how could ffmpeg deal with the newly created image ?

  The other option would be to get the 2 pics and play with them with openCv too, but then, hardcode a stream over, UDP in RTP style, but wouldnt be hard to display the stream nicely on the phone ?, and how to cut the image to fit in small packets ?

  Thanks for your help, I am still digging in the subject, so if you have any other way to do it I am taking !

  ps : My introducing Hello, won’t stay at the beginning of my message...
  Have a nice day every body.
  d-4 before frieday.

• convert avi to mp4 - centos - propper command

  16 novembre 2012, par Rares Daniel

  i need some help, please. Im strugling for a time now, with ffmpeg/mencoder/mp4box.

  i need to convert videos from avi to mp4, to maintain same quality, and IF possible, to reduce size.

  Im using centos 6.3 64bit

  i tried several commands :
  ffmpeg -i input.avi -acodec libfaac -b:a 128k -vcodec mpeg4 -b:v 1200k -flags +aic+mv4 output.mp4
  with : qscale 1

  the commands work, but the output file cant be played via httpd, i get (media time unsuported) or the output file is too large.
  I need an command to convert from avi to mp4 using h264 and faac.

  With windows , i have this "HandBrakeCLI.exe" -i "input" -t 1 -c 1 -o "output" -f mp4 —strict-anamorphic -e x264 -q 20 —vfr -a 1 -E faac -B 160 -6 dpl2 -R Auto -D 0 —gain=0 —audio-copy-mask none —audio-fallback ffac3 -x ref=1:weightp=1:subq=2:rc-lookahead=10:trellis=0:8x8dct=0 —verbose=1 And works like a charm. But not on linux

  Can you please give me a command to do this ? Also, if its possbile to use 2-3 core of server and 2-3 gb ram so the process would be faster.

  Thank you