[lmi-commits] [4771] Improve documentation

[Greg Chicares]

Revision: 4771
          http://svn.sv.gnu.org/viewvc/?view=rev&root=lmi&revision=4771
Author:   chicares
Date:     2010-02-10 14:44:36 +0000 (Wed, 10 Feb 2010)
Log Message:
-----------
Improve documentation

Modified Paths:
--------------
    lmi/trunk/ChangeLog
    lmi/trunk/fenv_lmi_x86.hpp

Modified: lmi/trunk/ChangeLog
===================================================================
--- lmi/trunk/ChangeLog 2010-02-10 14:43:52 UTC (rev 4770)
+++ lmi/trunk/ChangeLog 2010-02-10 14:44:36 UTC (rev 4771)
@@ -24324,3 +24324,16 @@
   [ChangeLog only]
 Fix defect introduced 20100127T1612Z.
 
+20100210T1443Z <address@hidden> [762]
+
+  DefectLog
+  alert.hpp
+  file_command_wx.cpp
+  materially_equal.hpp
+Provide direct links to usenet articles.
+
+20100210T1444Z <address@hidden> [762]
+
+  fenv_lmi_x86.hpp
+Improve documentation.
+

Modified: lmi/trunk/fenv_lmi_x86.hpp
===================================================================
--- lmi/trunk/fenv_lmi_x86.hpp  2010-02-10 14:43:52 UTC (rev 4770)
+++ lmi/trunk/fenv_lmi_x86.hpp  2010-02-10 14:44:36 UTC (rev 4771)
@@ -39,7 +39,7 @@
 /// is that C was strangely slow to adopt sophisticated numerics:
 ///   1980: 8087 introduced
 ///   1985: IEEE 754
-///   1999: C finally gets fesetenv(), but without precision control
+///   1999: C finally gets fesetenv(), but without precision control *
 /// which left setting the control word to nonstandard routines or
 /// nonportable asm, and part is that function _control87() in the
 /// widely-used ms C rtl takes an argument that differs gratuitously
@@ -57,6 +57,18 @@
 ///   RC rounding control
 /// Bit 12 is IC, the infinity-control bit, which is meaningful only
 /// for the 8087 and 80287, but settable (and ignored) on later chips.
+///
+/// * "without precision control"
+/// According to Goldberg:
+///   http://docs.sun.com/source/806-3568/ncg_goldberg.html#4130
+/// "fegetprec and fesetprec functions" were recommended in early
+/// drafts, but "this recommendation was removed before the changes
+/// were made to the C99 standard." The Rationale:
+///   www.open-std.org/jtc1/sc22/wg14/www/C99RationaleV5.10.pdf
+/// suggests that IEC 60559 is "ambivalent" as to whether precision
+/// control must be dynamic. Yet IEEE 754r Draft 1.2.5 [G.2] says:
+/// "changing the rounding direction or precision during execution may
+/// help identify subprograms that are unusually sensitive to roundoff"
 
 namespace fenv_lmi_x86 {} // doxygen workaround.