[m-dev.] Recent MR_sprintf_float fix and testcase output

[Keri Harris]

I have a couple of 10.04 tests failing on linux-ppc using the hlc.grade:

$ diff hard_coded/write_xml.exp2 hard_coded/write_xml.out
28c28
< 				<Float type="float">1.23553225220000e-97</Float>
---
> 				<Float type="float">1.2355322522e-97</Float>
407
< 							<Float 
type="float">1.23553225220000e-97</Float>
---
> 							<Float 
type="float">1.2355322522e-97</Float>


$ diff hard_coded/write_reg1.exp3 hard_coded/write_reg1.out
26,27c26,27
< 1.12832498300000e-21
< 2.23954899000000e+23
---
> 1.128324983e-21
> 2.23954899e+23


This looks to be a consequence of fixing a bug in mercury_float.c:

	revision 1.6
	runtime/mercury_float.c:
	    Fix a bug in MR_sprintf_float where it would convert 1.8e-10
	    into "1.80000000000000e-1" instead of "1.8e-10".

It may be worthwhile having a look through all the .exp[N] testcase files for 
other offending float outputs.


Thanks

Keri

--------------------------------------------------------------------------
mercury-developers mailing list
Post messages to:       mercury-developers at csse.unimelb.edu.au
Administrative Queries: owner-mercury-developers at csse.unimelb.edu.au
Subscriptions:          mercury-developers-request at csse.unimelb.edu.au
--------------------------------------------------------------------------