lua/testes, branch v5.5.1A mirror of https://github.com/lua/lua
https://git.lua4.win/lua/atom?h=v5.5.12026-07-23T16:58:30+00:00Small change in scope of variables in repeat-until2026-07-23T16:58:30+00:00Roberto Iroberto@inf.puc-rio.br2026-07-23T16:58:30+00:00urn:sha1:7579fc9d7ed90240487251dfb69168f8e64e9294
A close instruction is still inside the scope of the variables it is
closing. The extra close in a repeat-until (to close variables before
repeating the loop) was being coded outside that scope.
Bug: Issues with write barrier for __newindex2026-07-12T17:57:55+00:00Roberto Iroberto@inf.puc-rio.br2026-07-12T17:57:55+00:00urn:sha1:b996f8fd1be7fb711cc6f754a31a1c87d2c2fd9b
In 'luaV_finishset', there is an update on a table that is a field on
another table. If the first table is the same as the one with the field
(e.g., after 't.__newindex = t'), the update can change the value on
that field (e.g., there may be a collision and the field is moved, or
the field being updated is '__newindex' itself). After that, the
barrier is called with the table stored in that field, which is not
the correct table anymore.
Bug: 'lua_load' does not preserve the stack2026-04-23T20:57:42+00:00Roberto Iroberto@inf.puc-rio.br2026-04-23T20:57:42+00:00urn:sha1:3228a97c6a953dcf397944161bb64b12f1ff5384
'lua_load' does not preserve the stack through the calls to the
reader function, as it should. Immediately after the first call (to
detect whether chunk is binary) it adds stuff, and it also adds a new
table when starting the compilation of each new function.
Better error messages for vararg-table fields2026-04-13T17:06:23+00:00Roberto Iroberto@inf.puc-rio.br2026-04-13T17:06:23+00:00urn:sha1:d0bd25d2e7fb393a6d0a73645a099f9c3b9cc0a8Bug: wrong initialization in result from 'gmatch'2026-04-01T18:01:58+00:00Roberto Ierusalimschyroberto@inf.puc-rio.br2026-04-01T18:01:58+00:00urn:sha1:efddc2309c5ff8a1842bea8a9c0d7d4a5d6e1e60
Function returned by 'string.gmatch' can be left in an inconsistent
state after an error.
'table.tunpack' using 'aux_getn' like the others2026-03-09T19:24:49+00:00Roberto Iroberto@inf.puc-rio.br2026-03-09T19:24:49+00:00urn:sha1:377cbea61b2688b21c7d243fc0f42498851df794
'table.tunpack' was not checking its first argument, which could
result in error messages generated inside the API, without location
information.
BUG: shift overflow in utf-8 decode2026-02-18T16:24:04+00:00Roberto Iroberto@inf.puc-rio.br2026-02-18T16:24:04+00:00urn:sha1:10eb89d1141dc528806b32401e408e36fb2f3bf5
An initial byte \xFF will ask for 7 continuation bytes, and then the
shift by (count * 5) will try to shift 35 bits.
Environment variable for readline library name2026-01-30T19:47:33+00:00Roberto Iroberto@inf.puc-rio.br2026-01-30T19:47:33+00:00urn:sha1:c6b484823806e08e1756b1a6066a3ace6f080fae
The name of the readline library can be changed from its default value
through environment variable LUA_READLINELIB.
More effort in avoiding errors in finalizers2026-01-11T18:36:03+00:00Roberto Iroberto@inf.puc-rio.br2026-01-11T18:36:03+00:00urn:sha1:2a7cf4f319fc276f4554a8f6364e6b1ba4eb2ded
Before calling a finalizer, Lua not only checks stack limits, but
actually ensures that a minimum number of slots are already allocated
for the call. (If it cannot ensure that, it postpones the finalizer.)
That avoids finalizers not running due to memory errors that the
programmer cannot control.
Functions with vararg tables don't need hidden args.2025-11-28T18:12:51+00:00Roberto Iroberto@inf.puc-rio.br2025-11-28T18:12:51+00:00urn:sha1:a07f6a824197d7dc01c321599d3bc71936a2590e
Vararg functions with vararg tables don't use the arguments hidden in
the stack; therfore, it doesn't need to build/keep them.