On Apr 11, 2012, at 4:52 PM, Mike Stump wrote: > On Apr 11, 2012, at 7:04 AM, Bernd Schmidt wrote: >> On 04/11/2012 04:02 PM, Tristan Gingold wrote: >>> >>> On Apr 11, 2012, at 3:55 PM, Bernd Schmidt wrote: >>>> I'm working on a target where intptr_t and pointers are larger than >>>> size_t and ptrdiff_t. The testsuite has problems in this area, since we >>>> often use the latter two types for casting from/to pointers, leading to >>>> unwanted warnings. In some cases I've checked the corresponding PRs and >>>> found that the original testcases used something like plain int or long. >>> >>> Would this target be able to host gcc ? >> >> I do not wish to think about this :) It's embedded, so hopefully no one >> will try. >> >> In principle, I'd think any target with enough memory can be made to >> host gcc, with varying amounts of work. > > I've hosted gcc on mine, mainly as a code generation correctness check... > It's nice having a simulator and enough address bits, you can just plop down > yet more memory and presto, everything just works. The more annoying aspect > is having to wire up stat and readdir for sim. Fork and exec, are more > trivial by comparison. Some gcc ports might not have enough address bits, > that might be the only hard requirement that one cannot fudge.
VMS is an host/target where pointers (64 bits) can be larger than size_t and ptrdiff_t. We can run gcc on it, but we also still have a very few patches to submit to achieve that! Tristan.
