Jan Kundrát

ASAN/UBSAN GCC11 test failures (`executable_dll_and_plugin*`)

3

On the latest `dev` (30c3efdc9477080dc86fdaf717c05622ff6ca332), some tests (`executable_dll_and_plugin*`) fail when built with ASAN+UBSAN on GCC11 (NixOS). The exact-same configuration works fine when building without `-fsanitize=address,undefined`. Interestingly, stuff works with ASAN/UBSAN...

I'm working on multi-camare remote live production using all sorts of Canon EOS cameras, including the 5Div and RP models. These have a pretty decent AF implementation that's suitable for...

Since Python 3.10, the `asyncio.get_event_loop()` has been deprecated in favor of `asyncio.get_running_loop()`. That one, however, issues a warning when there's no event loop running (such as in this test suite)....

AddressSanitizer: heap-use-after-free via Netopeer2's test suite

3

I'm on the latest `devel` of everything, plus #2851: - CESNET/libyang@ba8cae2e - CESNET/libnetconf2@4d9d799 - `sysrepo` 5ec23c35d801740ce991e167104628e89f5d992c (`devel` plus #2851) - CESNET/Netopeer2@d6dc2f4 ``` 8/20 Test #8: test_subscribe_param .............Subprocess aborted***Exception: 19.20 sec...

CI: TSAN builds

6

There was an ASAN+UBSAN build job, but it was running on an old clang, and UBSAN failures were being silently ignored. Make sure that we're on the latest & greatest...

XPath: deref(): improve error messages

5

While the error message was accurate, it was not as helpful as it could be. Before: ``` libyang warn: Argument #1 of xpath_deref is a list node "amplifier". ``` Now:...

A TL;DR version: can you please report all ``must`` failures when validating data? It seems that only the first one of them is reported by ``yanglint`` and I presume also...

Support linking to issues, stories, epics and anything just by their number

2

I would like to have a uniform way of linking to bugs, user stories, epics, etc, of a given project hosted at tree.taiga.io just by their number. Right now the...

Today, it might be difficult to represent a network as-is in GNPy, that is, to completely switch off any autodesign-driven transformations. Here's a YANG proposal which (I hope) will make...

Today, fiber loss is specified in the IO format indirectly, via the `length` and the `loss_coef`. Vinay raised a good point in that users typically have access to OTDR measurements,...