go-algorand 3.0.1-stable #2949
Merged
go-algorand 3.0.1-stable #2949
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Treat LogCalls similarly to Eval Deltas with respect to encoding limits.
This PR addresses several different behaviors of pingpong that doesn't align (anymore) with it's usage:
1. during asset testing, avoid creating assets for the source account - this would generate failures on subsequent iterations.
2. for asset testing, ensure we have sufficient asset supply for the entire rebalancing period.
3. refactor the waiting for next round using a single `waitForNextRoundOrSleep` method.
4. generate an error when the number of created assets per account surpass the expected value.
5. when distributing assets, avoid using the `pps.constructTxn` method, as it's randomly select source/destination. instead, assemble the transaction explicitly.
6. `sendAsGroup` was much simplified by passing the explicit account. The non-possible error cases were removed.
7. in `prepareApps`, applications are compiled only once.
8. the`senders` slice in `prepareApps` should have been reset. It wouldn't work correctly without it.
9. The pingpong logic of sendFromTo was changed -
instead of sending traffic, rebalancing and sending traffic again, we calculate the pessimistic remaining balance on each of the accounts. A background go-routine would update all these account. Once the balance on 20% of the accounts drops below the desired level, we attempt to rebalance them. In addition, in 50% of the cases, the pingpong would pick an account that has a low balance as the recipient of the transaction.
This model allow much longer runs without rebalancing actions takes place, resulting in a more precise and consistent transaction generation.
10. precise rate control. The rate control was improved to ensure we generate transactions at a constant rate, with a 20ms wiggling temporal buffer.
11. a `constructPayment` was created to replace the libgoal `ConstructPayment`. The difference is in the source of the getSuggestedParams`, which are refreshed by the monitoring go-routine. This reduces unneeded communication between pingpong and the node.
12. fix some of the allocated funding balances for applications - making sure that we account for the setup costs, so that we won't run out of funds between rebalancing.
fixing a typo in e2e.sh for integration tests
## Summary
With this PR, the commands `goal app info`, `goal account info`, and `goal account list --info` now display the number of extra pages an application has.
For example, the following were executed on a private network where apps 1, 2, 3, 4 have 1, 2, 3, 0 extra pages, respectively.
```
$ goal app info --app-id 3 -d DATA_DIR
Application ID: 3
Creator: U4SKWGT6KYRA3DCMVPJLMMMNCP35UOYBFMQLEHLYDHO4C4N7GU7LLNQUHU
Approval hash: Q7ODX32ILHMGO6KPSELVTO7RCVOUZ37RWYHISAWER7GVARJGKTWEWD37AM
Clear hash: P7GEWDXXW5IONRW6XRIRVPJCT2XXEQGOBGG65VJPBUOYZEJCBZWTPHS3VQ
Extra program pages: 3
Max global byteslices: 2
Max global integers: 1
Max local byteslices: 4
Max local integers: 3
```
```
$ goal account info -a U4SKWGT6KYRA3DCMVPJLMMMNCP35UOYBFMQLEHLYDHO4C4N7GU7LLNQUHU -d DATA_DIR
Created Assets:
<none>
Held Assets:
<none>
Created Apps:
ID 1, 1 extra page, global state used 0/1 uints, 0/2 byte slices
ID 2, 2 extra pages, global state used 0/1 uints, 0/2 byte slices
ID 3, 3 extra pages, global state used 0/1 uints, 0/2 byte slices
ID 4, global state used 0/1 uints, 0/2 byte slices
Opted In Apps:
<none>
```
```
$ goal account list --info -d DATA_DIR
[offline] EHF6QYU3HCUUXIZAMPJ5RZWNDT37Y72NHVZ4XHWMYILPXWVU7G2G4PX32I EHF6QYU3HCUUXIZAMPJ5RZWNDT37Y72NHVZ4XHWMYILPXWVU7G2G4PX32I 4000024000000000 microAlgos
Created Assets:
...
[offline] U4SKWGT6KYRA3DCMVPJLMMMNCP35UOYBFMQLEHLYDHO4C4N7GU7LLNQUHU U4SKWGT6KYRA3DCMVPJLMMMNCP35UOYBFMQLEHLYDHO4C4N7GU7LLNQUHU 1000006000004997 microAlgos [created app IDs: 1, 2, 3, 4]
Created Assets:
<none>
Held Assets:
<none>
Created Apps:
ID 1, 1 extra page, global state used 0/1 uints, 0/2 byte slices
ID 2, 2 extra pages, global state used 0/1 uints, 0/2 byte slices
ID 3, 3 extra pages, global state used 0/1 uints, 0/2 byte slices
ID 4, global state used 0/1 uints, 0/2 byte slices
Opted In Apps:
<none>
```
## Test Plan
Modified existing e2e test and expect test. (I know expect tests aren't currently running, but it passed locally.)
<!-- Thanks for submitting a pull request! We appreciate the time and effort you spent to get this far. If you haven't already, please make sure that you've reviewed the CONTRIBUTING guide: https://github.com/algorand/go-algorand/blob/master/CONTRIBUTING.md#code-guidelines In particular ensure that you've run the following: * make generate * make sanity (which runs make fmt, make lint, make fix and make vet) It is also a good idea to run tests: * make test * make integration --> ## Summary A user reported a strange and not very helpful error message. This was due to incomplete error handling in the bash script relating to a truncated `updater` binary. The current workaround for this is to delete the `updater` binary and trigger `update.sh` to re-download it, but this is not obvious and I wouldn't expect any user to reach that conclusion. This fix adds error handling in case `updater` fails and should fix this particular failure. Here is the error message: ``` Current Version = 8590458880 Latest Version = ./update.sh: line 250: [: 8590458880: unary operator expected New version found Update Downloaded to /tmp/tmp.QlV71NLGFr/.tar.gz Expanding update... tar (child): /tmp/tmp.QlV71NLGFr/.tar.gz: Cannot open: No such file or directory tar (child): Error is not recoverable: exiting now tar: Child returned status 2 tar: Error is not recoverable: exiting now *** UPDATE FAILED: Error expanding update *** ``` ## Test Plan Ran some tests against a directory, empty file, and regular file in my shell: ``` will@arendelle:~/go/src/github.com/algorand/go-algorand/cmd/updater$ [[ -f dir && -s dir ]] will@arendelle:~/go/src/github.com/algorand/go-algorand/cmd/updater$ echo $? 1 will@arendelle:~/go/src/github.com/algorand/go-algorand/cmd/updater$ [[ -f empty && -s empty ]] will@arendelle:~/go/src/github.com/algorand/go-algorand/cmd/updater$ echo $? 1 will@arendelle:~/go/src/github.com/algorand/go-algorand/cmd/updater$ [[ -f util.go && -s util.go ]] will@arendelle:~/go/src/github.com/algorand/go-algorand/cmd/updater$ echo $? 0 ```
Codegen verification step is now being run by CircleCI, so it's no longer necessary to run it on Travis, especially since we recently got a `toomanyrequests: You have reached your pull rate limit.` docker error. We want to reduce unnecessary docker requests.
parametrize account balance range for bootstrapping testing scenario.
ParseHostOrURL does not accepts ":xxx:123" anymore pingpongTest.exp is still disabled
Add comments for protocol hash IDs and tags
The number of buckets in verifiedTransactionCache can be arbitrarily large depending on the requested cache size, and each lookup in the cache does that many map lookups. This PR limits the number of buckets to 3. Now, the lookup takes at most 3 lookups and the size overhead is 1.5.
When creating large networks with bootstrapped databases ( 60m accounts ), the netgoal was failing due to a false-positive deadlock detection. Given that the netgoal is a testing utility, it would be better to turn off these false-positives testing off netgoal.
Status code differs on CircleCI. Relaxed the status code value check. Change the skipped expect tests to be defined as "t.Skip" so we can track these as such on the log file.
fix e2e test TestStartAndStop
…2784) This also unifies a lot of field handling so that the code is less susceptible to copy-paste errors, but it has a ways to go. Hard to both unify the handling of the different fields and use separate types for each field index. Lots and lots of interfaces could do. Or generics in 6 months? For the moment, each field type is a pile of arrays and maps working together. It's unclear what we gain from those separate index types, since we always cast into the type just before using them, as we are getting the indexes from reading a byte in bytecode. New tests will catch this mistake if we make it again.
* After re-enabling we started seeing failures on other e2e-go tests due to unsufficient resources to allow algod making new rounds * Possible solution is to run expect tests as a separate stage on CI go test ./test/e2e-go/cli/goal/expect/... -run TestGoalWithExpect
Disassemble was omitting labels if they were only used as targets of backjumps. This fixes that, but making a second pass to find labels if any such labels exist in a program. It also changes testProg, which is used extensively to test programs in the logic package, so that every such test is a Disassemble/Reassemble test to ensure the same program is produced.
new scripts runNodeHost.py and nodeHostTarget.py manage starting a new ec2 node with algod (or several) pointed at a specific relay so that we can test per-client usage on that one relay. improve goal: don't crash if kmd isn't present when we don't need it tweaks to other heapWatch scripts for measuring this data.
Attempt to split expect tests into a 3rd category. This changes the behavior of e2e_go_tests.sh. Previously it would run the e2e_go tests AND the EXPECT tests. Now it runs one or the other.
This PR adds a Cost Parameter to DryrunResponse. It is calculated by giving the evalParams object in dryrun.go the maximum budget allowed (maxappcost * maxgroupsize) and checking the difference after stateful evaluation. LogicSig dryruns will return cost of 0.
Fix overriding consensus parameters in evaluator for indexer
Goal was not able to communicate with the Ledger Nano S device. The device has two HID occurrences on the OS and after the firmware upgrade, one of them became not responsive. That made goal to the hang. fixing the interaction with the Ledger devices (S and X) on the Windows platform.
…stable2.10.1-remerge
Relstable2.10.1 remerge
* expose nonparticipation txn field in teal
* Increase max app opt ins to 50 * Fixed error check in new ledger testing harness
Allow AVM 1.0 programs to "execute transactions" 1. An app now has authority over an account which is the hash of its appid. 2. The app can create transactions with three new opcodes, tx_begin, tx_field, and tx_submit. 3. If the transaction so created would need to be authorized by that app account from 1, the effects occur. 4. Fees on inner transactions are paid from the app account, though overpaying at the top-level can be pooled. So far, only pay and axfer are allowed. acfg and afrz are under consideration, but have separate issues written. This makes ApplyData a recursive datastructure, so that the "inner" transactions of app calls can be encoded in EvalDelta and, themselves, have ApplyData.
…2840) Intended to just increase coverage in AVM, but doing so found a bug. (Go figure.) tx_field should only allow actual addresses for address fields, rather than low integers as references from the Account array.
* Add txnas opcodes * Add args opcode
Drive the checking of whether tx_field can set a particular field in a table-driven way, rather than by simple switch. This is cleaner as we add functionality in later versions, and will allow generating the spec for tx_field.
Our partitioned tests might be skipping tests or running them twice. Here we want to make sure they don't do that.
loads and stores opcodes that take scratch slot from stack
When multiple threads use the synchTest object, and one fails, all call FailNow. After the first call, the main test routine reads t.finished to terminate the test. While the main test routine is reading t.finished, another thread calls FailNow, which in turn writes to t.finished, causing the data race. While FailNow called from synchTest are guarded by a mutex, the main testing object has the TestingTB object, which does not use the same mutex. The solution is to call FailNow from synchTest object only the first time. * When multiple threads use the synchTest object, and one fails, all call FailNow. After the first call, the main test routine reads t.finished to terminate the test. While the main test routine is reading t.finished, another thread calls FailNow, which in turn writes to t.finished, causing the data race. While FailNow called from synchTest are guarded by a mutex, the main testing object has the TestingTB object, which does not use the same mutex. The solution is to call FailNow from synchTest object only the first time.
* Add secp256k1 module from go-ethereum 1.10.7 * ECDSA secp256k1 implementation Cost calculated relative to ed25519verify BenchmarkEcDsa/ecdsa_verify-8 17341 67335 ns/op BenchmarkEcDsa/ecdsa_pk_decompress-8 42854 25714 ns/op BenchmarkEcDsa/ecdsa_pk_recover-8 15337 78249 ns/op BenchmarkEd25519Verifyx1-8 15244 75253 ns/op
Indexer needs to know all accounts referenced in a block to be able to read all of them in one batch. This PR creates a function that returns all such account addresses.
The auto-generated msgp test files were not up-to-date. This PR updates these.
Updated the netgoal utility to only create one wallet per non-participating node to better concentrate Algos for use by pingpong.
Increment TxnCounter with inner transactions acfg and afrz inner transactions This also sorts out inner txcounting, so that an innner transaction knows the current txncount as it executes, so it can get the right id if it creates. Co-authored-by: Jason Paulos <[email protected]>
Rename extractXXbits, add small spec updates.
Since the merge of #2852 go mod tidy and go mod vendor see paths in github.com/ethereum/go-ethereum/crypto/secp256k1/libsecp256k1 as a dependency, and go mod tidy adds it to go.mod.
This commit makes consensus parameter changes for V30. v30 introduces AVM 1.0 and TEAL 5, increases the app opt in limit to 50, and allows costs to be pooled in grouped stateful transactions.
go-algorand 3.0.0-beta
 # Highlights This version adds debugging for itxn programs and includes a bugfix for dryrun crashes. **This release does not contain a protocol upgrade.** # Changes 1. TEAL * Added * Enable debugging itxn programs (#2900) * Fixed * Fix dryrun crash on rewards calculation (#2894) ## Additional Resources * [Algorand Forum](https://forum.algorand.org) * [Developer Documentation](https://developer.algorand.org)
Codecov Report
@@ Coverage Diff @@
## rel/stable #2949 +/- ##
==============================================
+ Coverage 47.10% 47.32% +0.21%
==============================================
Files 349 355 +6
Lines 56351 57182 +831
==============================================
+ Hits 26544 27059 +515
- Misses 26831 27069 +238
- Partials 2976 3054 +78
Continue to review full report at Codecov.
|
algojack
requested review from
bricerisingalgorand,
algobarb,
egieseke,
onetechnical and
tsachiherman
algobarb
approved these changes
Sep 24, 2021
PhearZero
pushed a commit
to PhearNet/crypto
that referenced
this pull request
Jan 17, 2025
go-algorand 3.0.1-stable
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This release requires a protocol upgrade.
This release contains a consensus protocol upgrade, which implements the following spec: https://github.com/algorandfoundation/specs/tree/bc36005dbd776e6d1eaf0c560619bb183215645c
Highlights
This release features the 1.0 version of AVM, extending Algorand's smart contract capabilities for even greater composability. Highlights in this release include:
Changes
Changed
goal app info& show logs are binary in REST API (Display app address ingoal app info& show logs are binary in REST API #2836)Fixed
startEvaluator()(SimplifystartEvaluator()#2812)verifiedTransactionCacheto 3 (Limit the number of buckets inverifiedTransactionCacheto 3 #2775)Additional Resources