#3260: Linking stage2 on PPC gives "scattered reloc r_address too large"
---------------------------------+------------------------------------------
Reporter: benl | Owner:
Type: bug | Status: new
Priority: normal | Milestone: 6.14.1
Component: Compiler | Version: 6.13
Keywords: | Testcase:
Blockedby: | Difficulty: Unknown
Os: MacOS X | Blocking:
Architecture: Unknown/Multiple | Failure: None/Unknown
---------------------------------+------------------------------------------
Changes (by mokus):
* cc: mo...@… (added)
Comment:
For what it's worth, I have run into this bug now when compiling my own
code (on Mac OS 10.5 and 10.6 with GHC 6.12.3).
The code that tripped it is a TH-generated library of data types and
binary encoding/decoding functions for a massive object model I have no
control over. Just the enumerations in this data model are big enough to
trip it. They constitute approximately 400 types, each with instances of
Eq, Ord, Show, Enum, Bounded, and Binary. Just the 'Types' module itself
which contains nothing but the data declarations with derived Show
instances has a 21M object file which, if I understand what I'm reading
here correctly, is already over the size limit of what LD can handle.
Once this ticket is resolved, am I correct in understanding that to make
things like this work I will still have to split these declarations across
multiple modules each compiling to a small-enough object file? That's
easy enough in the case of enumerations, but the other data types in this
model have a big ad-hoc web of interdependencies that would make the task
somewhat difficult to do automatically (although there are only about 100
of those so maybe I'd be OK).
Anyway, just wanted to make it known that this problem has impacted user
code, not just GHC itself.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/3260#comment:19>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
