lua-users home
lua-l archive

[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]


Hi,

In Lua 5.4.4, the C functions "luai_makeseed" (in lstate.c) and "l_randomizePivot" (in ltablib.c) both seem to be trying to introduce some "unsigned int" sized entropy into Lua, both using "time(NULL)" (among other sources of entropy), and both working around uncertainty in the size of a "time_t" value.  They are also both protected by the PreProcessor directives (I guess) to allow alternative definitions.  Curiously, "luai_makeseed" also uses the value of "L" and the address of the "lua_newstate" function to add entropy, while the "l_randomizePivot" uses the the c function "clock".  There is a similar function "randseed" (in "lmathlib.c"), though it is a little bit different, not being PreProcessor protected and attempting to generate more bits entropy, but from only "time(NULL)" and the value of "L".

My Use-Case is an attempt to port Lua on to a Raspberry Pi Pico microcontroller - a rather limited environment in which the C "time" function (and "clock" function) returns the error code -1 as allowed in the C standard (see https://en.cppreference.com/w/c/chrono/time).  On the other hand, like other microcontrollers, it has other sources of entropy.  In my case, there is a definite source of "random" bits, but on other Microcontrollers, one might try using the low order bits of some analogue signal input.

I know that randomness is a Huge topic, but my point is perhaps a bit simpler.  Could we tweak the Lua implementation to offer a single PreProcessor protected source of "unsigned int" sized entropy, that by default could use "time(NULL)" (among others) but for which ports to other platforms might provide an alternative source of entropy.  This function would replace "luai_makeseed" and "l_randomisePivot" and could be used by "randseed" instead of "time(NULL)".  It would also mean that there would be no loss of entropy if the C functions supported on the platform only returned the error code -1.

My unlikely worst-case-scenario is that lua's table.sort function is called, during which it repeatedly calls "l_randomizePivot" to provide alternative pivot points, but which keep returning the same value because "time" and "clock" are both returning -1.

Cheers,

Steve