Closed wurli closed 8 months ago
Thank you! I will merge it, but not right now because we are still finding bugs in code that replicate Nvim-R
's features. For example, I just noted that when I reboot the computer and start to edit a Quarto file with auto_start="always"
, Valgrind reports many errors. The first one is:
==4616== Memcheck, a memory error detector
==4616== Copyright (C) 2002-2022, and GNU GPL'd, by Julian Seward et al.
==4616== Using Valgrind-3.20.0 and LibVEX; rerun with -h for copyright info
==4616== Command: /home/aquino/software/R-4.3.3/library/nvimcom/bin/rnvimserver
==4616== Parent PID: 4584
==4616==
==4616== Invalid write of size 8
==4616== at 0x10D161: load_pkg_data (data_structures.c:642)
==4616== by 0x10D41F: finish_bol (data_structures.c:711)
==4616== by 0x10D535: finished_building_objls (data_structures.c:737)
==4616== by 0x10EC90: stdin_loop (rnvimserver.c:161)
==4616== by 0x10EF80: main (rnvimserver.c:253)
==4616== Address 0x4abec20 is 32 bytes inside a block of size 80 free'd
==4616== at 0x48431EF: free (in /usr/libexec/valgrind/vgpreload_memcheck-amd64-linux.so)
==4616== by 0x10C4C3: pkg_delete (data_structures.c:247)
==4616== by 0x10CAEF: update_pkg_list (data_structures.c:397)
==4616== by 0x10F01A: ParseMsg (tcp.c:61)
==4616== by 0x10F816: get_whole_msg (tcp.c:328)
==4616== by 0x10F891: receive_msg (tcp.c:348)
==4616== by 0x48F845B: start_thread (pthread_create.c:444)
==4616== by 0x4978ADF: clone (clone.S:100)
This is going to be slow to fix because I have to reboot my laptop to replicate it (or to confirm that it's fixed). I will barely need to use R during the next weeks. So, it would be great if more people could run the rnvimserver
through Valgrind to find bugs as well as reliable ways of replicating them.
It seems that I fixed the bug visible in Valgrind log...
This should hopefully draw a little more attention to the note that Nvim-R is superseded for Neovim users.