libcwal  Check-in [71969d4b29]

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
SHA1:71969d4b29f8962a694355b4666fcb8784492445
Date: 2016-02-16 10:45:06
User: stephan
Comment:oops - enabled one of the #if's which allows th1ish to continue to work (it was disabled while testing yesterday).
Tags And Properties
  • branch=th1ish-R.I.P. inherited from [18c978ea9c]
  • sym-th1ish-R.I.P. inherited from [18c978ea9c]
Context
2016-02-16
10:45
oops - enabled one of the #if's which allows th1ish to continue to work (it was disabled while testing yesterday). Leaf check-in: [71969d4b29] user: stephan tags: th1ish-R.I.P.
09:32
Moving th1ish out of trunk, as it's holding up a couple pending improvements in the core (and hasn't been hacked on in over 18 months other than to accommodations a few minor API changes). Its sloppy ref handling (by current standards) makes it untenable in conjunction with up-coming cwal improvements. th1ish is now dead. Long live th1ish (or not -... check-in: [18c978ea9c] user: stephan tags: th1ish-R.I.P.
Changes

Changes to cwal.c.

3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
           e.a.value.b = e.b;
           e.b.value.a = e.a
        */
        if(CWAL_RCFLAG_HAS(w,CWAL_RCF_IS_DESTRUCTING)){
            return;
        }
        assert( CWAL_REFCOUNT(w) || V_IS_BUILTIN(w) );
#if 0
/*20160215: only needed for th1ish, not s2!*/
        /*
           This cwal_unref_from() doesn't _seem_ to necessary, but i'm
           not quite sure why. Ah, _possibly_ because the downstream
           code is doing everything exactly right vis-a-vis refs and
           upscoping?








|







3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
           e.a.value.b = e.b;
           e.b.value.a = e.a
        */
        if(CWAL_RCFLAG_HAS(w,CWAL_RCF_IS_DESTRUCTING)){
            return;
        }
        assert( CWAL_REFCOUNT(w) || V_IS_BUILTIN(w) );
#if 1
/*20160215: only needed for th1ish, not s2!*/
        /*
           This cwal_unref_from() doesn't _seem_ to necessary, but i'm
           not quite sure why. Ah, _possibly_ because the downstream
           code is doing everything exactly right vis-a-vis refs and
           upscoping?