Hi Frank



 FA> I also believe that in a multitask environment, that would be

 FA> fatal if another program got a slice between free() and return,



 PC> not sure if i understand what you mean here?.



 FA> Well i imagined a scenario, where Prog1 is running in conjunction

 FA> with Prog2, Prog1 goes off and does it's thing in foo() and just as it

 FA> executed the free(str) statement, OS steps in and gives Prog2 a slice

 FA> Prog2 happens to be looking for memory and comes across a
"free" block

 FA> where Prog1's str lives, and happily populates it...



 FA> When Prog1 gets it's slice back it goes and returns ptr which is no

 FA> longer the same animal,



ptr is the same, it's just the block of memory that it points to

that may have changed. And since your code has already freed that

block, you shouldn't expect its contents to be still valid. For

example, you might have a function call between the free() and the

return that might call an alloc type function. Or an interrupt

routine could pop its head up just then and do the same.



 FA> I also believe that in a multitask environment, that would be

 FA> fatal if another program got a slice between free() and return,



Nothing actually wrong with returning ptr after it's been freed,

it's just pointless. Unless it's used to return an error if

something in the function failed (a bit obfuscated for my taste,

unless fully commented...)



Cheers



--- PPoint 1.88