Fixing run time exception: reference into empty vector #4968

pull ENikS wants to merge 1 commits into bitcoin:master from ENikS:vector_subscript_out_of_range changing 2 files +7 −7

ENikS commented at 5:35 PM on September 23, 2014: contributor

Replacing it with more predictable behavior.
Fixing reference into empty vector run time exception. Replacing it with more predictable behavior. 744f7c6522
BitcoinPullTester commented at 5:49 PM on September 23, 2014: none

Automatic sanity-testing: PASSED, see http://jenkins.bluematt.me/pull-tester/p4968_744f7c6522c23e635a32ba5c327a0b62d49ef6ee/ for binaries and test log. This test script verifies pulls every time they are updated. It, however, dies sometimes and fails to test properly. If you are waiting on a test, please check timestamps to verify that the test.log is moving at http://jenkins.bluematt.me/pull-tester/current/ Contact BlueMatt on freenode if something looks broken.
sipa commented at 5:51 PM on September 23, 2014: member

ACK
in src/test/base58_tests.cpp:None in 744f7c6522
```
  39 | @@ -40,7 +40,7 @@ BOOST_AUTO_TEST_CASE(base58_EncodeBase58)
  40 |          std::vector<unsigned char> sourcedata = ParseHex(test[0].get_str());
  41 |          std::string base58string = test[1].get_str();
  42 |          BOOST_CHECK_MESSAGE(
  43 | -                    EncodeBase58(&sourcedata[0], &sourcedata[sourcedata.size()]) == base58string,
  44 | +                    EncodeBase58(begin_ptr(sourcedata), begin_ptr(sourcedata) + sourcedata.size()) == base58string,
```
laanwj commented at 9:37 AM on September 25, 2014:

Nit: begin_ptr(sourcedata) + sourcedata.size() could be end_ptr(sourcedata)

ENikS commented at 2:56 PM on September 25, 2014:

One of these 'DA' moments. Should have thought that if there is a begin_ptr, there should be end_ptr as well.

Perhaps a pointer to useful macros should go to a coding guidelines as well?

laanwj commented at 3:25 PM on September 25, 2014:

Indeed, that could be useful to add.
ENikS commented at 3:35 PM on September 25, 2014:

Not to split hairs but would it be a bit more appropriate to use

&v.front() instead of &v[0]

Not that it matters, but since we are so big on style...
laanwj commented at 8:19 AM on September 26, 2014:

Believe me, we're not really big on style :) We're somewhat insistent on not changing files to a meltingpot of different code styles, but apart from that, I always prefer pull requests that actually fix an issue (as you've been doing).
laanwj referenced this in commit b50e5fe099 on Sep 25, 2014
laanwj commented at 10:47 AM on September 25, 2014: member

Merged via b50e5fe (with nit fix)
laanwj closed this on Sep 25, 2014

in src/script/interpreter.cpp:None in 744f7c6522

 653 | +                        CHash256().Write(begin_ptr(vch), vch.size()).Finalize(begin_ptr(vchHash));
 654 |                      popstack(stack);
 655 |                      stack.push_back(vchHash);
 656 |                  }
 657 | -                break;
 658 | +                break;

Diapolo commented at 11:15 AM on September 25, 2014:

@laanwj Why are whitespace changes acceptable here?

laanwj commented at 11:19 AM on September 25, 2014:

They are not; but I must have missed this, and it seems accidental anyway.

ENikS deleted the branch on Sep 25, 2014
MarcoFalke locked this on Sep 8, 2021

Contributors

Linked (view graph)

#4970 Debug depends/stl