Craig A. Berry wrote:
> When hv.c was in the midst of being refactored recently, I would
> occasionally see a null pointer passed as the first argument to
> setenv(). The final version doesn't do that but it made me realize a
> decent setenv() implementation should handle the null pointer but our
> homegrown one was not. The attached patch makes it do so.

Thanks, applied as #21864.

Reply via email to