On 2020/01/05 07:43, Nazar Zhuk wrote: > On 2020-01-04 09:21, Stuart Henderson wrote: > > On 2020-01-04, Nazar Zhuk <[email protected]> wrote: > > > I get SCRIPT_FILENAME passed from httpd relative to httpd chroot > > > (/site1/htdocs/... ) and PHP being chrooted into /var/www/site1 needs > > > that to be relative to it's own chroot (/htdocs/...). > > > > httpd is a bit inflexible (intentionally, I think). Can you work around > > it with ln -s . /var/www/site1/site1 ? > > Yes, this works, thanks. > > Is this intentional though, or is it that this use case hasn't been fully > considered yet?
Unsure. > Separate chroots for different FastCGI processes seems like a good idea to > isolate those processes from each other. > > This could be configured with something like: > > fastcgi strip <n> > > similar to how "request strip" works. > > I can write a patch if this is something the maintainers would be interested > in adding. It does sound like it might be useful and should be fairly straightforward - I thinik it would make sense to write a patch and send to tech@ and see what httpd users/devs think.
