RFR: 8335906: [s390x] Test Failure: GTestWrapper.java

[Amit Kumar]

On Tue, 9 Jul 2024 05:51:47 GMT, Thomas Stuefe <stuefe at openjdk.org> wrote:

>> PAT_1 is fine because only 1 element is there, i.e. last element: 
>> 
>> [debug] u.c[idx] =   6a,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  j value =   6a, v =   6a 
>> [debug] u.c[idx] =   64,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  d value =   64, v =   64 
>> [debug] u.c[idx] =   6b,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  k value =   6b, v =   6b 
>> [debug] u.c[idx] =   2f,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  / value =   2f, v =   2f 
>> [debug] u.c[idx] =   69,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  i value =   69, v =   69 
>> [debug] u.c[idx] =   6e,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  n value =   6e, v =   6e 
>> [debug] u.c[idx] =   74,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  t value =   74, v =   74 
>> [debug] u.c[idx] =   65,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  e value =   65, v =   65 
>> [debug] u.c[idx] =   72,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  r value =   72, v =   72 
>> [debug] u.c[idx] =   6e,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  n value =   6e, v =   6e 
>> [debug] u.c[idx] =   61,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  a value =   61, v =   61 
>> [debug] u.c[idx] =   6c,  sizeof(u.v) = 8, unitsize = 1 , i+1 = 1,  l value =   6c, v =   6c 
>> 
>> 
>> issue starts when array `u.c` size increases; 
>> 
>> PAT_2:
>> [debug] u.c[idx] =   64,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 1,  d value = 6a64, v = 6a64 
>> [debug] u.c[idx] =   6a,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 2,  j value = 6a64, v = 6a64 
>> [debug] u.c[idx] =   2f,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 1,  / value = 6b2f, v = 6b2f 
>> [debug] u.c[idx] =   6b,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 2,  k value = 6b2f, v = 6b2f 
>> [debug] u.c[idx] =   6e,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 1,  n value = 696e, v = 696e 
>> [debug] u.c[idx] =   69,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 2,  i value = 696e, v = 696e 
>> [debug] u.c[idx] =   65,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 1,  e value = 7465, v = 7465 
>> [debug] u.c[idx] =   74,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 2,  t value = 7465, v = 7465 
>> [debug] u.c[idx] =   6e,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 1,  n value = 726e, v = 726e 
>> [debug] u.c[idx] =   72,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 2,  r value = 726e, v = 726e 
>> [debug] u.c[idx] =   6c,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 1,  l value = 616c, v = 616c 
>> [debug] u.c[idx] =   61,  sizeof(u.v) = 8, unitsize = 2 , i+1 = 2,  a value = 616c, v = 616c 
>> [debug] u.c[idx] =   6c,  sizeof(u.v) = 8, unitsize = 2...
>
> Looks good, Amit. Many thanks for fixing this.

There is no regression on s390x (tier1) test. @tstuefe should we consider this as trivial fix or wait for another approval ?

-------------

PR Comment: https://git.openjdk.org/jdk/pull/20089#issuecomment-2216699184