yes I meant "volatile" not "virtual".
I should not drink and post.
Michael David Crawford, Baritone
mdcrawford@xxxxxxxxx
One Must Not Trifle With Wizards For It Makes Us Soggy And Hard To Light.
On Thu, Feb 25, 2016 at 9:20 AM, Ingo Weinhold <ingo_weinhold@xxxxxx> wrote:
On 25.02.2016 17:42, Michael David Crawford wrote:
I'm working at silencing build warnings. I figure this is a good way
to become familiar with the code base.
GCC warned that a variable "might be clobbered by 'longjmp' or
'vfork'". That's the first time I've ever seen that. The problem has
to do with that registers are restored by the setjmp that occurs after
longjmp.
It worked to declare that variable virtual. Somehow I find that
unsatisfying. My objective in declaring it virtual was to get the
variable stored on the stack rather than in a register.
I assume you mean "volatile".
I agree that this is not a satisfying solution, though I don't have a better
solution. In the JPEGTranslator case one would think that rephrasing the
code "if (setjmp(...) != 0) return ...;" should do the trick, since the
compiler could infer that the variable won't be used after a longjmp(), but
at least gcc 2 doesn't.
There are more warnings just like this in PNGTranslator.c. When I
declare the variables virtual, it leads to fatal errors because some
of these variables are passed to subroutines whose arguments are not
declared virtual.
I could declare the arguments virtual but again this strikes me as
unsatisfying. Is there something better I could do?
Assuming that those are pointer variables, you might have declared them
incorrectly. E.g. "volatile int* foo" declares the location the pointer
refers to volatile, while "int* volatile foo" would declare the pointer
variable itself volatile. Only the latter can be assigned to "int*".
CU, Ingo
