🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: gnu.gcc.bug
Article #2731

inline suggestion

#2731
From: Jim Meyering
Date: Tue, 23 May 1989 19:57
23 lines
1385 bytes
A suggestion for inline_able-function evaluation criterion:

   Upon several occasions, I have given functions the inline attribute
   and have been informed that they are too large to be made inline.
   However, if they had been made inline, the combination of copy
   propagation and dead code elimination would have reduced the size
   of the resulting version of the function by a factor as large as six
   in one case.

   If it is essential that explicitly declared inline functions not be
   larger than some particular size, is it possible to calculate code size
   after dead code elimination?

   More realistically, might it be possible to manually override the
   default limit on the size of functions that may be made inline?
   Or maybe, impose no size restriction for functions that are explicitly
   declared inline?

   Some good examples are the manually unrolled BLAS from LINPACK,
   especially scopy().

---
Jim Meyering      meyering@cs.utexas.edu
Technical Details
Message-ID: <8905231957.AA19167@ai.cs.utexas.edu>
Path: arc1-ca.pugleaf.net!utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!CS.UTEXAS.EDU!meyering

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!CS.UTEXAS.EDU!meyering
From: meyering@CS.UTEXAS.EDU (Jim Meyering)
Newsgroups: gnu.gcc.bug
Subject: inline suggestion
Message-ID: <8905231957.AA19167@ai.cs.utexas.edu>
Date: 23 May 89 19:57:40 GMT
Sender: daemon@tut.cis.ohio-state.edu
Distribution: gnu
Organization: UofTX @ Austin CS Department
Lines: 23