arm - correctly handle denormal results during softfp subtraction

2018-08-02  Nicolas Pitre <nico@fluxnic.net>

	PR libgcc/86512
	* config/arm/ieee754-df.S (adddf3): Don't shortcut denormal handling
	when exponent goes negative. Update my email address.
	* config/arm/ieee754-sf.S (addsf3): Likewise.

From-SVN: r263267

libgcc/ChangeLog

+2018-08-02  Nicolas Pitre <nico@fluxnic.net>
+
+	PR libgcc/86512
+	* config/arm/ieee754-df.S: Don't shortcut denormal handling when
+	exponent goes negative. Update my email address.
+	* config/arm/ieee754-sf.S: Likewise.
+
 2018-08-01  Martin Liska  <mliska@suse.cz>
 
 	* libgcov-profiler.c (__gcov_indirect_call_profiler_v2): Do not

libgcc/config/arm/ieee754-df.S

 /* ieee754-df.S double-precision floating point support for ARM
 
    Copyright (C) 2003-2018 Free Software Foundation, Inc.
-   Contributed by Nicolas Pitre (nico@cam.org)
+   Contributed by Nicolas Pitre (nico@fluxnic.net)
 
    This file is free software; you can redistribute it and/or modify it
    under the terms of the GNU General Public License as published by the
@@ -238,9 +238,10 @@ LSYM(Lad_a):
 	movs	ip, ip, lsl #1
 	adcs	xl, xl, xl
 	adc	xh, xh, xh
-	tst	xh, #0x00100000
-	sub	r4, r4, #1
-	bne	LSYM(Lad_e)
+	subs	r4, r4, #1
+	do_it	hs
+	cmphs	xh, #0x00100000
+	bhs	LSYM(Lad_e)
 
 	@ No rounding necessary since ip will always be 0 at this point.
 LSYM(Lad_l):

libgcc/config/arm/ieee754-sf.S

 /* ieee754-sf.S single-precision floating point support for ARM
 
    Copyright (C) 2003-2018 Free Software Foundation, Inc.
-   Contributed by Nicolas Pitre (nico@cam.org)
+   Contributed by Nicolas Pitre (nico@fluxnic.net)
 
    This file is free software; you can redistribute it and/or modify it
    under the terms of the GNU General Public License as published by the
@@ -168,9 +168,10 @@ LSYM(Lad_e):
 LSYM(Lad_a):
 	movs	r1, r1, lsl #1
 	adc	r0, r0, r0
-	tst	r0, #0x00800000
-	sub	r2, r2, #1
-	bne	LSYM(Lad_e)
+	subs	r2, r2, #1
+	do_it	hs
+	cmphs	r0, #0x00800000
+	bhs	LSYM(Lad_e)
 
 	@ No rounding necessary since r1 will always be 0 at this point.
 LSYM(Lad_l):