API tools faq
paste
Guest User

Clang produces a much worse binary than GCC

a guest
Jun 29th, 2026
106
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Bash 6.15 KB | None | 0 0
raw download clone embed print report
  1. GCC 16.1.1 O3 dynamic generic:
  2. ~/Games/3DS roms
  3. ❯ ./benchmark_script.sh Pokemon\ X\ \(USA\)\ \(En\,Ja\,Fr\,De\,Es\,It\,Ko\).zcci
  4.  
  5.  Performance counter stats for process id '1361':
  6.  
  7.                  0      context-switches:u               #      0.0 cs/sec  cs_per_second
  8.                  0      cpu-migrations:u                 #      0.0 migrations/sec  migrations_per_second
  9.                 11      page-faults:u                    #      1.5 faults/sec  page_faults_per_second
  10.           7,161.70 msec task-clock:u                     #      0.3 CPUs  CPUs_utilized
  11.        193,566,172      L1-dcache-load-misses:u          #      2.7 %  l1d_miss_rate            (8.95%)
  12.        193,518,378      L1-icache-load-misses:u          #      1.2 %  l1i_miss_rate            (9.03%)
  13.        115,637,336      LLC-loads:u                      #      7.3 %  llc_miss_rate            (9.21%)
  14.         77,430,567      branch-misses:u                  #      2.3 %  branch_miss_rate         (8.97%)
  15.      3,266,708,109      branches:u                       #    456.1 M/sec  branch_frequency     (9.19%)
  16.     22,212,794,142      cpu-cycles:u                     #      3.1 GHz  cycles_frequency       (18.13%)
  17.     20,803,875,357      instructions:u                   #      0.9 instructions  insn_per_cycle  (17.84%)
  18.      6,865,549,781      dTLB-loads:u                     #      0.2 %  dtlb_miss_rate           (9.31%)
  19.     20,742,011,698      iTLB-loads:u                     #      0.0 %  itlb_miss_rate           (9.13%)
  20.  
  21.       25.003241657 seconds time elapsed
  22.  
  23.  
  24. ~/Games/3DS roms took 40s
  25. ❯ agamemodeauto:
  26. ./benchmark_script.sh Pokemon\ X\ \(USA\)\ \(En\,Ja\,Fr\,De\,Es\,It\,Ko\).zcci
  27.  
  28.  Performance counter stats for process id '1713':
  29.  
  30.                  0      context-switches:u               #      0.0 cs/sec  cs_per_second
  31.                  0      cpu-migrations:u                 #      0.0 migrations/sec  migrations_per_second
  32.                 12      page-faults:u                    #      1.7 faults/sec  page_faults_per_second
  33.           7,004.27 msec task-clock:u                     #      0.3 CPUs  CPUs_utilized
  34.        173,728,982      L1-dcache-load-misses:u          #      2.6 %  l1d_miss_rate            (9.24%)
  35.        212,435,290      L1-icache-load-misses:u          #      1.4 %  l1i_miss_rate            (9.21%)
  36.        109,528,440      LLC-loads:u                      #      8.2 %  llc_miss_rate            (9.30%)
  37.         79,356,888      branch-misses:u                  #      2.5 %  branch_miss_rate         (9.18%)
  38.      3,439,686,220      branches:u                       #    491.1 M/sec  branch_frequency     (8.99%)
  39.     21,741,041,307      cpu-cycles:u                     #      3.1 GHz  cycles_frequency       (18.01%)
  40.     22,058,046,715      instructions:u                   #      1.0 instructions  insn_per_cycle  (18.01%)
  41.      6,771,307,635      dTLB-loads:u                     #      0.2 %  dtlb_miss_rate           (8.98%)
  42.     19,638,332,577      iTLB-loads:u                     #      0.0 %  itlb_miss_rate           (9.00%)
  43.  
  44.       25.003582007 seconds time elapsed
  45.  
  46. Clang 21.1.8 O3 flto generic:
  47. ~/Games/3DS roms
  48. ❯ ./benchmark_script.sh Pokemon\ X\ \(USA\)\ \(En\,Ja\,Fr\,De\,Es\,It\,Ko\).zcci
  49.  
  50.  Performance counter stats for process id '2180':
  51.  
  52.                  0      context-switches:u               #      0.0 cs/sec  cs_per_second
  53.                  0      cpu-migrations:u                 #      0.0 migrations/sec  migrations_per_second
  54.                 13      page-faults:u                    #      0.9 faults/sec  page_faults_per_second
  55.          13,726.07 msec task-clock:u                     #      0.5 CPUs  CPUs_utilized
  56.      2,117,259,283      L1-dcache-load-misses:u          #     25.8 %  l1d_miss_rate            (9.16%)
  57.        184,058,651      L1-icache-load-misses:u          #      0.5 %  l1i_miss_rate            (9.04%)
  58.        712,281,103      LLC-loads:u                      #      3.4 %  llc_miss_rate            (9.00%)
  59.         71,955,056      branch-misses:u                  #      2.3 %  branch_miss_rate         (9.24%)
  60.      3,167,229,978      branches:u                       #    230.7 M/sec  branch_frequency     (9.10%)
  61.     43,574,971,572      cpu-cycles:u                     #      3.2 GHz  cycles_frequency       (18.15%)
  62.     26,612,017,585      instructions:u                   #      0.6 instructions  insn_per_cycle  (18.21%)
  63.      8,118,270,936      dTLB-loads:u                     #      0.4 %  dtlb_miss_rate           (9.03%)
  64.     26,826,570,433      iTLB-loads:u                     #      0.0 %  itlb_miss_rate           (9.01%)
  65.  
  66.       25.002604861 seconds time elapsed
  67.  
  68.  
  69. ~/Games/3DS roms took 40s
  70. ❯ gamemodeauto:
  71. ./benchmark_script.sh Pokemon\ X\ \(USA\)\ \(En\,Ja\,Fr\,De\,Es\,It\,Ko\).zcci
  72.  
  73.  Performance counter stats for process id '2419':
  74.  
  75.                  0      context-switches:u               #      0.0 cs/sec  cs_per_second
  76.                  0      cpu-migrations:u                 #      0.0 migrations/sec  migrations_per_second
  77.                 14      page-faults:u                    #      1.0 faults/sec  page_faults_per_second
  78.          13,716.60 msec task-clock:u                     #      0.5 CPUs  CPUs_utilized
  79.      2,143,125,658      L1-dcache-load-misses:u          #     26.5 %  l1d_miss_rate            (9.13%)
  80.        176,731,420      L1-icache-load-misses:u          #      0.5 %  l1i_miss_rate            (9.03%)
  81.        706,377,730      LLC-loads:u                      #      3.5 %  llc_miss_rate            (9.02%)
  82.         70,778,866      branch-misses:u                  #      2.2 %  branch_miss_rate         (9.43%)
  83.      3,094,315,317      branches:u                       #    225.6 M/sec  branch_frequency     (9.01%)
  84.     43,477,039,997      cpu-cycles:u                     #      3.2 GHz  cycles_frequency       (18.14%)
  85.     26,503,573,521      instructions:u                   #      0.6 instructions  insn_per_cycle  (18.15%)
  86.      8,167,911,244      dTLB-loads:u                     #      0.4 %  dtlb_miss_rate           (9.00%)
  87.     26,439,996,980      iTLB-loads:u                     #      0.0 %  itlb_miss_rate           (9.03%)
  88.  
  89.       25.002684910 seconds time elapsed
Tags: benchmark performance clang LLVM azahar
Advertisement
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!