Closed div72 closed 4 months ago
Absolutely. If this works and the tests pass without the fudge function, we should remove the fudging to restore the sensitivity.The other stuff that was failing in the wallet tests I was able to fix by using the Allocation class and avoiding doubles altogether.Sent from my iPhoneOn Mar 2, 2024, at 10:07 AM, div72 @.***> wrote:See the commit message on the first commit for more details.
@jamescowens I reverted your fixes since I think it's better if we get tests failures in case our assumptions are violated in the future in a similar way, but I can drop the reversions if you want.
You can view, comment on, or merge this pull request online at: https://github.com/gridcoin-community/Gridcoin-Research/pull/2746
Commit Summary
f80976d build: enforce SSE2 on x86 targets feabb82 Revert "Implement comp_double comparison function in certain tests" 4228fed Revert "Cleanup wallet_tests.cpp and remove the rest of double usage" 069d1e0 Revert "Change some tests to use Fraction class in wallet tests"
File Changes (4 files)
M
configure.ac
(10)
M
src/test/gridcoin/claim_tests.cpp
(34)
M
src/test/gridcoin/researcher_tests.cpp
(68)
M
src/test/wallet_tests.cpp
(150)
Patch Links:
https://github.com/gridcoin-community/Gridcoin-Research/pull/2746.patch https://github.com/gridcoin-community/Gridcoin-Research/pull/2746.diff
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: @.***>
Ah. I realized you reverted all of the commits. I think we should keep the fixes in wallet tests (not the others) as fp should probably never have been used there in the first place?
@theMarix as far as I am concerned this puts the nail in the coffin on this issue. The gcc13 compiler on OpenSUSE tumbleweed 32-bit is broken and we should file a bug report. See @div72's comments in https://github.com/gridcoin-community/Gridcoin-Research/pull/2746/commits/f80976d0e29ba0d62541c6cc5d3cb533b2e2d2c6.
See the commit message on the first commit for more details.
@jamescowens I reverted your fixes since I think it's better if we get tests failures in case our assumptions are violated in the future in a similar way, but I can drop the reversions if you want.