13:29 <pepijndevos_> welp, getting some weird issues... maybe uninitialised memory or some pointer that got into a IdString :(

13:30 <daveshah> valgrind it

13:31 <pepijndevos_> how?

13:31 <pepijndevos_> also lol:

13:31 <pepijndevos_> Info: port cpu.cpuregs.regs[11][20]_DFFE_Q_DFFLC.I2, connected to net 'cpu.reg_pc[22]_LUT4_I1_I2[2]', has timing budget of 340282346638528859811704183484516925440.000000ns

13:31 <daveshah> usually valgrind will tell you if you are relying on uninitialised memory

13:31 <pepijndevos_> that's 0xffffff00000000000000000000000000 in hex

13:34 <pepijndevos_> usually you have to run valgrind with a particular tool right

13:35 <daveshah> i just the default one for this kind of stuff

13:35 <pepijndevos_> alright

13:35 <daveshah> it will spew a load of crap at the start from python, you can either ignore that or find the flag to not print stuff coming from that so

13:35 <daveshah> or build without python

13:39 <pepijndevos_> trying...

13:41 <pepijndevos_> I forgot how slow valgrind is...

13:49 <pepijndevos_> seeing lots of ==25727== Invalid read of size 4

13:49 <pepijndevos_> ==25727== at 0x1866B0: nextpnr_gowin::Context::checksum() const (nextpnr.cc:366)

13:49 <pepijndevos_> but seems irrelevant

13:50 <pepijndevos_> ==25727== Invalid read of size 8

13:50 <pepijndevos_> ==25727== at 0x17DCF2: std::_Hashtable<nextpnr_gowin::IdString, std::pair<nextpnr_gowin::IdString const, nextpnr_gowin::PortInfo>, std::allocator<std::pair<nextpnr_gowin::IdString const, nextpnr_gowin::PortInfo> >, std::__detail::_Select1st, std::equal_to<nextpnr_gowin::IdString>, std::hash<nextpnr_gowin::IdString>, std::__detail::_Mod_range_hashing, std::__detail::_Default_ranged_hash, std::__detail::_Prime_rehash_policy, std::__detail::_Hashta

13:50 <pepijndevos_> ble_traits<false, false, true> >::_M_find_before_node(unsigned long, nextpnr_gowin::IdString const&, unsigned long) const (hashtable.h:1573)

13:50 <pepijndevos_> enjoy your C++ soup

13:50 <pepijndevos_> lemme google what that even means

13:51 <daveshah> looks like a bad cell port to me

13:52 <daveshah> incidentally, it probably won't help here but if you are debugging a packer I'd strongly recommend having https://github.com/YosysHQ/nextpnr/commit/70de8b3a03a62cd1be8732cf5916800322a12a17 in your tree if it isn't already

13:53 <pepijndevos_> it is not, thanks

13:54 <pepijndevos_> yea, probably my initialization is all wrong...

14:13 <daveshah> I think what is probably happening is a reference to a deleted cell is remaining on a net

16:33 <pepijndevos_> hmmmm okay, so that could be for example packing deleting a cell, but missing one of its inputs?

16:37 <pepijndevos_> Yea what's probably happening... the Gowin DFF cells have LSR and CE, but the generic DFF does not, so I forgot to deal with those inputs.