Skip to contentSkip to conversation

lib/tests: Remove tst_fuzzy_sync02

PATCH New Petr Vorel 4 participants 5 messages Aug 31, 2021

lore
Share & tools
lore

Petr Vorel

author
#0 lore raw

PATCH 1/1 lib/tests: Remove tst_fuzzy_sync02

tst_fuzzy_sync02 is not stable enough [1]:

../../include/tst_fuzzy_sync.h:346: TINFO: loop = 10001, delay_bias = 3960

../../include/tst_fuzzy_sync.h:335: TINFO: start_a - start_b: { avg = -8859ns, avg_dev =    66ns, dev_ratio = 0.01 }
../../include/tst_fuzzy_sync.h:335: TINFO: end_a - start_a  : { avg = 17160ns, avg_dev =    85ns, dev_ratio = 0.00 }
../../include/tst_fuzzy_sync.h:335: TINFO: end_b - start_b  : { avg = 18474ns, avg_dev =    55ns, dev_ratio = 0.00 }
../../include/tst_fuzzy_sync.h:335: TINFO: end_a - end_b    : { avg = -10173ns, avg_dev =   124ns, dev_ratio = 0.01 }
../../include/tst_fuzzy_sync.h:335: TINFO: spins            : { avg =  2751  , avg_dev =    31  , dev_ratio = 0.01 }

../../include/tst_fuzzy_sync.h:685: TINFO: Exceeded execution loops, requesting exit tst_fuzzy_sync02.c:169: TFAIL: 2| =:48
...
runtest TINFO: === C TEST RESULTS ===
...
runtest TINFO: 1x TFAIL: tst_fuzzy_sync02

Both tst_fuzzy_sync0{1,2} occasionally fails (restart helps). But tst_fuzzy_sync03 is stable thus we have at least smoke test for fuzzy sync.

[1] https://github.com/linux-test-project/ltp/runs/3471271675

---
 lib/newlib_tests/runtest.sh | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
lib/newlib_tests/runtest.sh +2 -2
index 70fbffb6c..05c76228b 100755
@@ -2,8 +2,8 @@
 # Copyright (c) 2021 Petr Vorel <[email protected]>
 
 LTP_C_API_TESTS="${LTP_C_API_TESTS:-test05 test07 test09 test12 test15 test18
-tst_bool_expr test_exec test_timer tst_res_hexd tst_strstatus tst_fuzzy_sync02
-tst_fuzzy_sync03 test_zero_hugepage.sh}"
+tst_bool_expr test_exec test_timer tst_res_hexd tst_strstatus tst_fuzzy_sync03
+test_zero_hugepage.sh}"
 
 LTP_SHELL_API_TESTS="${LTP_SHELL_API_TESTS:-shell/tst_check_driver.sh shell/net/*.sh}"
 
Signature · 1 line
2.33.0
#2 lore raw

PATCH 1/1 lib/tests: Remove tst_fuzzy_sync02

In reply to Petr Vorel

Hello Petr,

Quoted 15 lines · Petr Vorel

tst_fuzzy_sync02 is not stable enough [1]:

../../include/tst_fuzzy_sync.h:346: TINFO: loop = 10001, delay_bias = 3960

../../include/tst_fuzzy_sync.h:335: TINFO: start_a - start_b: { avg = -8859ns, avg_dev =    66ns, dev_ratio = 0.01 }
../../include/tst_fuzzy_sync.h:335: TINFO: end_a - start_a  : { avg = 17160ns, avg_dev =    85ns, dev_ratio = 0.00 }
../../include/tst_fuzzy_sync.h:335: TINFO: end_b - start_b  : { avg = 18474ns, avg_dev =    55ns, dev_ratio = 0.00 }
../../include/tst_fuzzy_sync.h:335: TINFO: end_a - end_b    : { avg = -10173ns, avg_dev =   124ns, dev_ratio = 0.01 }
../../include/tst_fuzzy_sync.h:335: TINFO: spins            : { avg =  2751  , avg_dev =    31  , dev_ratio = 0.01 }

../../include/tst_fuzzy_sync.h:685: TINFO: Exceeded execution loops, requesting exit tst_fuzzy_sync02.c:169: TFAIL: 2| =:48
...
runtest TINFO: === C TEST RESULTS ===
...
runtest TINFO: 1x TFAIL: tst_fuzzy_sync02

Quoted 3 lines

Both tst_fuzzy_sync0{1,2} occasionally fails (restart helps). But tst_fuzzy_sync03 is stable thus we have at least smoke test for fuzzy sync.

Yes, there is probably some work to do on the algorithm itself. At least to make it stable in this kind of environment.

Signature · 2 lines
Thank you,
Richard.
#3 lore raw

PATCH 1/1 lib/tests: Remove tst_fuzzy_sync02

In reply to Petr Vorel

Hi Petr,

I have no objection to disable it too.

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20210831/fb25f87a/attachment-0001.htm>
author
#4 lore raw

PATCH 1/1 lib/tests: Remove tst_fuzzy_sync02

Hi all,

thanks, merged!

Kind regards,
Petr