add payload attestation pool

[Tomi-3-0]

No description provided.

[tersec]

Can run make test locally to regenerate test summaries to fix the Jenkins CI errors, which do exactly this and verify that the test summaries are current/consistent:

[2025-11-14T03:34:35.919Z] + git diff --exit-code --ignore-submodules=all
[2025-11-14T03:34:35.919Z] diff --git a/AllTests-mainnet.md b/AllTests-mainnet.md
[2025-11-14T03:34:35.919Z] index 1fc63000f..74a9b36cb 100644
[2025-11-14T03:34:35.919Z] --- a/AllTests-mainnet.md
[2025-11-14T03:34:35.919Z] +++ b/AllTests-mainnet.md
[2025-11-14T03:34:35.919Z] @@ -836,6 +836,15 @@ AllTests-mainnet
[2025-11-14T03:34:35.919Z]  ```diff
[2025-11-14T03:34:35.919Z]  + pre-1.1.0                                                                                  OK
[2025-11-14T03:34:35.919Z]  ```
[2025-11-14T03:34:35.919Z] +## Payload attestation pool [Preset: mainnet]
[2025-11-14T03:34:35.919Z] +```diff
[2025-11-14T03:34:35.919Z] ++ Can add and retrieve payload attestations [Preset: mainnet]                                OK
[2025-11-14T03:34:35.919Z] ++ Can get payload attestations for block production [Preset: mainnet]                        OK
[2025-11-14T03:34:35.919Z] ++ Different payload presence values [Preset: mainnet]                                        OK
[2025-11-14T03:34:35.919Z] ++ Duplicate validator in PTC - multiple signatures [Preset: mainnet]                         OK
[2025-11-14T03:34:35.919Z] ++ Multiple validators in PTC can attest [Preset: mainnet]                                    OK
[2025-11-14T03:34:35.919Z] ++ Payload attestations get pruned [Preset: mainnet]                                          OK
[2025-11-14T03:34:35.919Z] +```
[2025-11-14T03:34:35.919Z]  ## PeerPool testing suite
[2025-11-14T03:34:35.919Z]  ```diff
[2025-11-14T03:34:35.919Z]  + Access peers by key test

[tersec]

This can be an iterator and avoid allocating the seq[int], because it's only used by

        for ptc_index in ptc_positions:
          aggregation_bits[ptc_index] = true
          signatures.add(cookedSig)

Instead of positions.add(i), it would yield i.

[github-actions]

Unit Test Results

       12 files  ±  0    2 436 suites  +4   48m 27s ⏱️ +42s
12 649 tests +  6  12 084 ✔️ +  6  565 💤 ±0  0 ❌ ±0 
63 612 runs  +24  62 884 ✔️ +24  728 💤 ±0  0 ❌ ±0 

Results for commit abf5e3b. ± Comparison against base commit 1c55749.

♻️ This comment has been updated with latest results.

[tersec]

The

gcc  @nimcache/release/all_tests/all_tests.makefile.linkerArgs
/github-runner/github-runner-node-02/workspace/nimbus-eth2/nimbus-eth2/beacon_chain/consensus_object_pools/payload_attestation_pool.nim: In function ‘_ZN24payload_attestation_pool17aggregateMessagesEN24payload_attestation_pool22PayloadAttestationPoolEN9constants4SlotE3ptrIN24payload_attestation_pool23PayloadAttestationEntryEE3varIN4base10StateCacheEE.isra’:
/github-runner/github-runner-node-02/workspace/nimbus-eth2/nimbus-eth2/beacon_chain/consensus_object_pools/payload_attestation_pool.nim:90:15: error: stack usage is 5664224 bytes [-Werror=stack-usage=]
   90 | proc aggregateMessages(
      |               ^
lto1: some warnings being treated as errors
make[2]: *** [/tmp/ccr4dSTb.mk:383: /tmp/ccUDECaB.ltrans127.ltrans.o] Error 1
make[2]: *** Waiting for unfinished jobs....
lto-wrapper: fatal error: make returned 2 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status
make[1]: *** [nimcache/release/all_tests/all_tests.makefile:3856: build] Error 1
make: *** [Makefile:373: all_tests] Error 2

errors are deliberate stack-usage-checking:

nimbus-eth2/config.nims

Lines 64 to 70 in 1c55749

	if defined(limitStackUsage):
	# This limits stack usage of each individual function to 1MB - the option is
	# available on some GCC versions but not all - run with `-d:limitStackUsage`
	# and look for .su files in "./build/", "./nimcache/" or $TMPDIR that list the
	# stack size of each function.
	switch("passC", "-fstack-usage -Werror=stack-usage=1048576")
	switch("passL", "-fstack-usage -Werror=stack-usage=1048576")

and appear to be due to:

• Stack materialization, zeroing, and copying of purely heap-based data with iterator and case object nim-lang/Nim#25287

[Tomi-3-0]

The

gcc  @nimcache/release/all_tests/all_tests.makefile.linkerArgs
/github-runner/github-runner-node-02/workspace/nimbus-eth2/nimbus-eth2/beacon_chain/consensus_object_pools/payload_attestation_pool.nim: In function ‘_ZN24payload_attestation_pool17aggregateMessagesEN24payload_attestation_pool22PayloadAttestationPoolEN9constants4SlotE3ptrIN24payload_attestation_pool23PayloadAttestationEntryEE3varIN4base10StateCacheEE.isra’:
/github-runner/github-runner-node-02/workspace/nimbus-eth2/nimbus-eth2/beacon_chain/consensus_object_pools/payload_attestation_pool.nim:90:15: error: stack usage is 5664224 bytes [-Werror=stack-usage=]
   90 | proc aggregateMessages(
      |               ^
lto1: some warnings being treated as errors
make[2]: *** [/tmp/ccr4dSTb.mk:383: /tmp/ccUDECaB.ltrans127.ltrans.o] Error 1
make[2]: *** Waiting for unfinished jobs....
lto-wrapper: fatal error: make returned 2 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status
make[1]: *** [nimcache/release/all_tests/all_tests.makefile:3856: build] Error 1
make: *** [Makefile:373: all_tests] Error 2

errors are deliberate stack-usage-checking:

nimbus-eth2/config.nims

Lines 64 to 70 in 1c55749

	if defined(limitStackUsage):
	# This limits stack usage of each individual function to 1MB - the option is
	# available on some GCC versions but not all - run with `-d:limitStackUsage`
	# and look for .su files in "./build/", "./nimcache/" or $TMPDIR that list the
	# stack size of each function.
	switch("passC", "-fstack-usage -Werror=stack-usage=1048576")
	switch("passL", "-fstack-usage -Werror=stack-usage=1048576")

and appear to be due to:

• Stack materialization, zeroing, and copying of purely heap-based data with iterator and case object nim-lang/Nim#25287

Thanks for the pointer! That explains exactly what I have been seeing.
Whenever I set my systems stack limit (with ulimit) low to ~8162KB I get seg faults and when I increase the stack limit the tests pass fine.

[tersec]

The

gcc  @nimcache/release/all_tests/all_tests.makefile.linkerArgs
/github-runner/github-runner-node-02/workspace/nimbus-eth2/nimbus-eth2/beacon_chain/consensus_object_pools/payload_attestation_pool.nim: In function ‘_ZN24payload_attestation_pool17aggregateMessagesEN24payload_attestation_pool22PayloadAttestationPoolEN9constants4SlotE3ptrIN24payload_attestation_pool23PayloadAttestationEntryEE3varIN4base10StateCacheEE.isra’:
/github-runner/github-runner-node-02/workspace/nimbus-eth2/nimbus-eth2/beacon_chain/consensus_object_pools/payload_attestation_pool.nim:90:15: error: stack usage is 5664224 bytes [-Werror=stack-usage=]
   90 | proc aggregateMessages(
      |               ^
lto1: some warnings being treated as errors
make[2]: *** [/tmp/ccr4dSTb.mk:383: /tmp/ccUDECaB.ltrans127.ltrans.o] Error 1
make[2]: *** Waiting for unfinished jobs....
lto-wrapper: fatal error: make returned 2 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status
make[1]: *** [nimcache/release/all_tests/all_tests.makefile:3856: build] Error 1
make: *** [Makefile:373: all_tests] Error 2

errors are deliberate stack-usage-checking:

nimbus-eth2/config.nims

Lines 64 to 70 in 1c55749

	if defined(limitStackUsage):
	# This limits stack usage of each individual function to 1MB - the option is
	# available on some GCC versions but not all - run with `-d:limitStackUsage`
	# and look for .su files in "./build/", "./nimcache/" or $TMPDIR that list the
	# stack size of each function.
	switch("passC", "-fstack-usage -Werror=stack-usage=1048576")
	switch("passL", "-fstack-usage -Werror=stack-usage=1048576")

and appear to be due to:

• Stack materialization, zeroing, and copying of purely heap-based data with iterator and case object nim-lang/Nim#25287

Thanks for the pointer! That explains exactly what I have been seeing. Whenever I set my systems stack limit (with ulimit) low to ~8162KB I get seg faults and when I increase the stack limit the tests pass fine.

Yes, the 5664224 bytes the error mentions is only for one function, not the whole stack, so it's plausible it would require 2MB+ more of stack for other functions.

[tersec]

If this is only useful for the tests, can move it into tests.

It's nice to have a static guarantee of that sort that the main code isn't using the inefficient approach.

[Tomi-3-0]

I use it here also:

nimbus-eth2/beacon_chain/consensus_object_pools/payload_attestation_pool.nim

Line 95 in abf5e3b

let ptc_list = get_ptc_list(forkyState.data, slot, cache)