You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi there, I’ve come across a notable regression issue in Julia 1.9.3. Would anyone happen to have any insights on this?
Interestingly, the script below doesn’t allocate any memory in Julia 1.8.1, but it leads to 200.00k allocations: 6.104 MiB in Julia 1.9.3.
This seems to be fixed in 1.10-beta3.
functionf(arr)
s =0for el in arr
if el ∈ ("1", "2")
s+=1endendreturn s
end
a =string.(rand(Int, 100000));
@timef(a) # Compilation consumes the majority of the time.@timef(a) # So, we need to run it a second time
I have investigated the problem and found that the reason for the allocations is the presence of ("1", "2"). To confirm this, you can run the second example.
functiong(arr, arg)
s =0for el in arr
if el ∈ arg
s+=1endendreturn s
end
a =string.(rand(Int, 100000));
b = ("1", "2")
@timeg(a, b)
@timeg(a, b) # In Julia 1.9.3, allocations will be reduced to 0.
The text was updated successfully, but these errors were encountered:
Hi there, I’ve come across a notable regression issue in Julia 1.9.3. Would anyone happen to have any insights on this?
Interestingly, the script below doesn’t allocate any memory in Julia 1.8.1, but it leads to 200.00k allocations: 6.104 MiB in Julia 1.9.3.
This seems to be fixed in 1.10-beta3.
I have investigated the problem and found that the reason for the allocations is the presence of ("1", "2"). To confirm this, you can run the second example.
The text was updated successfully, but these errors were encountered: