Recherche avancée

Sur d’autres sites (10392)

• 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.

• Announcement : Piwik to focus on Reliability, Performance and Security

  7 octobre 2014, par Matthieu Aubry — About, Community

  To our valued team and community,

  Well, we have moved fast and achieved so much during the past few months. Relentlessly releasing major version after major version… We got a lot done including several major new features !

  The speed of adding new features was a great showcase of how agile our small teams and the larger community are. And I’m so proud to see automated testing becoming common practice among everyone hacking on Piwik !

  For the next few months until the new year we will focus on making what we have better. We will fix those rare but longstanding critical bugs, and aim to solve all Major issues and other must-have performance and general improvements. The core team and Piwik PRO will have the vision of making the existing Piwik and all plugins very stable and risk free. This includes edge cases, general bugs but also specific performance issues for high traffic or issues with edge case data payloads.

  We’ll be more pro-active and take Piwik platform to the next level of Performance, Security, Privacy & Reliability ! We will prove to the world that Free/Libre Web software can be of the highest standard of quality. By focusing on quality we will make Piwik even easier to maintain and improve in the future. We are building the best open platform that will let every user liberate their data and keep full control of it.

  If you have any feedback or questions get in touch or let’s continue the discussion in the forum.

  Thank you for your trust and for liberating your data with Piwik,

  Matthieu Aubry
  Piwik founder

  More information

  • How long is each release of Piwik maintained for ? (FAQ)
  • When is the next release of Piwik ? What is the release schedule ? (FAQ)
  • Pulse of Piwik : our monthly pulse report shows 27 different people contributed to Piwik core in the last month !

  

  This is an amazing testament of the power of free/libre software and yet we think this is just the beginning. We hope more developers will join and contribute to the Piwik project !

• Announcement : Piwik to focus on Reliability, Performance and Security

  7 octobre 2014, par Matthieu Aubry — About, Community

  To our valued team and community,

  Well, we have moved fast and achieved so much during the past few months. Relentlessly releasing major version after major version… We got a lot done including several major new features !

  The speed of adding new features was a great showcase of how agile our small teams and the larger community are. And I’m so proud to see automated testing becoming common practice among everyone hacking on Piwik !

  For the next few months until the new year we will focus on making what we have better. We will fix those rare but longstanding critical bugs, and aim to solve all Major issues and other must-have performance and general improvements. The core team and Piwik PRO will have the vision of making the existing Piwik and all plugins very stable and risk free. This includes edge cases, general bugs but also specific performance issues for high traffic or issues with edge case data payloads.

  We’ll be more pro-active and take Piwik platform to the next level of Performance, Security, Privacy & Reliability ! We will prove to the world that Free/Libre Web software can be of the highest standard of quality. By focusing on quality we will make Piwik even easier to maintain and improve in the future. We are building the best open platform that will let every user liberate their data and keep full control of it.

  If you have any feedback or questions get in touch or let’s continue the discussion in the forum.

  Thank you for your trust and for liberating your data with Piwik,

  Matthieu Aubry
  Piwik founder

  More information

  • How long is each release of Piwik maintained for ? (FAQ)
  • When is the next release of Piwik ? What is the release schedule ? (FAQ)
  • Pulse of Piwik : our monthly pulse report shows 27 different people contributed to Piwik core in the last month !

  

  This is an amazing testament of the power of free/libre software and yet we think this is just the beginning. We hope more developers will join and contribute to the Piwik project !