Power10 stub selection

This patch better supports mixing of power10 and non-power10 code,
as might be seen in a cpu-optimized library using ifuncs to select
functions optimized for a given cpu.  Using -Wl,--no-power10-stubs
isn't that good in this situation since non-power10 notoc stubs are
slower and larger than the power10 variants, which you'd like to use
on power10 code paths.

With this change, power10 pc-relative code that makes calls marked
@notoc uses power10 stubs if stubs are necessary, and other calls use
non-power10 instructions in stubs.  This will mean that if gcc is
generating code for -mcpu=power10 but with pc-rel disabled then you'll
get the older stubs even on power10 (unless you force with
-Wl,--power10-stubs).  That shouldn't be too big a problem: stubs that
use r2 are reasonable.  It's just the ones that set up addressing
using "mflr 12; bcl 20,31,.+4; mflr 11; mtlr 12" that should be
avoided if possible.

bfd/
	* elf64-ppc.c (struct ppc_link_hash_table): Add has_power10_relocs.
	(select_alt_stub): New function.
	(ppc_get_stub_entry): Use it here.
	(ppc64_elf_check_relocs): Set had_power10_relocs rather than
	power10_stubs.
	(ppc64_elf_size_stubs): Clear power10_stubs here instead.  Don't
	merge notoc stubs with other varieties when power10_stubs is "auto".
	Instead dup the stub hash table entry.
	(plt_stub_size, ppc_build_one_stub, ppc_size_one_stub): Adjust
	tests of power10_stubs.
ld/
	* emultempl/ppc64elf.em (power10-stubs): Accept optional "auto" arg.
	* ld.texi (power10-stubs): Update.
	* testsuite/ld-powerpc/callstub-1.d: Force --power10-stubs.
	* testsuite/ld-powerpc/callstub-2.d: Relax branch offset comparison.
	* testsuite/ld-powerpc/callstub-4.d: New test.
	* testsuite/ld-powerpc/notoc.d: Force --no-power10-stubs.
	* testsuite/ld-powerpc/notoc3.d,
	* testsuite/ld-powerpc/notoc3.s,
	* testsuite/ld-powerpc/notoc3.wf: New test.
	* testsuite/ld-powerpc/powerpc.exp: Run new tests.  Pass
	--no-power10-stubs for notoc link.

11 files changed, 257 insertions, 22 deletions

diff --git a/ld/ChangeLog b/ld/ChangeLog
index 32b4616..51a4e9c 100644
--- a/ld/ChangeLog
+++ b/ld/ChangeLog
@@ -1,3 +1,17 @@
+2020-07-19  Alan Modra  <amodra@gmail.com>
+
+	* emultempl/ppc64elf.em (power10-stubs): Accept optional "auto" arg.
+	* ld.texi (power10-stubs): Update.
+	* testsuite/ld-powerpc/callstub-1.d: Force --power10-stubs.
+	* testsuite/ld-powerpc/callstub-2.d: Relax branch offset comparison.
+	* testsuite/ld-powerpc/callstub-4.d: New test.
+	* testsuite/ld-powerpc/notoc.d: Force --no-power10-stubs.
+	* testsuite/ld-powerpc/notoc3.d,
+	* testsuite/ld-powerpc/notoc3.s,
+	* testsuite/ld-powerpc/notoc3.wf: New test.
+	* testsuite/ld-powerpc/powerpc.exp: Run new tests.  Pass
+	--no-power10-stubs for notoc link.
+
 2020-07-17  Hans-Peter Nilsson  <hp@bitrange.com>
 
 	* scripttempt/mmo.sc: Move .init first in .text output section.
diff --git a/ld/emultempl/ppc64elf.em b/ld/emultempl/ppc64elf.em
index 4987243..1331d03 100644
--- a/ld/emultempl/ppc64elf.em
+++ b/ld/emultempl/ppc64elf.em
@@ -716,7 +716,7 @@ PARSE_AND_LIST_LONGOPTS=${PARSE_AND_LIST_LONGOPTS}'
   { "no-plt-align", no_argument, NULL, OPTION_NO_PLT_ALIGN },
   { "plt-localentry", optional_argument, NULL, OPTION_PLT_LOCALENTRY },
   { "no-plt-localentry", no_argument, NULL, OPTION_NO_PLT_LOCALENTRY },
-  { "power10-stubs", no_argument, NULL, OPTION_POWER10_STUBS },
+  { "power10-stubs", optional_argument, NULL, OPTION_POWER10_STUBS },
   { "no-power10-stubs", no_argument, NULL, OPTION_NO_POWER10_STUBS },
   { "emit-stub-syms", no_argument, NULL, OPTION_STUBSYMS },
   { "no-emit-stub-syms", no_argument, NULL, OPTION_NO_STUBSYMS },
@@ -773,7 +773,7 @@ PARSE_AND_LIST_OPTIONS=${PARSE_AND_LIST_OPTIONS}'
   --no-plt-localentry         Don'\''t optimize ELFv2 calls\n"
 		   ));
   fprintf (file, _("\
-  --power10-stubs             Use Power10 PLT call stubs (default auto)\n"
+  --power10-stubs [=auto]     Use Power10 PLT call stubs (default auto)\n"
 		   ));
   fprintf (file, _("\
   --no-power10-stubs          Don'\''t use Power10 PLT call stubs\n"
@@ -889,7 +889,20 @@ PARSE_AND_LIST_ARGS_CASES=${PARSE_AND_LIST_ARGS_CASES}'
       break;
 
     case OPTION_POWER10_STUBS:
-      params.power10_stubs = 1;
+      if (optarg != NULL)
+	{
+	  if (strcasecmp (optarg, "auto") == 0)
+	    params.power10_stubs = -1;
+	  else if (strcasecmp (optarg, "yes") == 0)
+	    params.power10_stubs = 1;
+	  else if (strcasecmp (optarg, "no") == 0)
+	    params.power10_stubs = 0;
+	  else
+	    einfo (_("%F%P: invalid --power10-stubs argument `%s'\''\n"),
+		   optarg);
+	}
+      else
+	params.power10_stubs = 1;
       break;
 
     case OPTION_NO_POWER10_STUBS:
diff --git a/ld/ld.texi b/ld/ld.texi
index 165a3d9..2a93e94 100644
--- a/ld/ld.texi
+++ b/ld/ld.texi
@@ -7922,15 +7922,13 @@ care.  @option{--no-plt-localentry} is the default.
 @itemx --no-power10-stubs
 When PowerPC64 @command{ld} links input object files containing
 relocations used on power10 prefixed instructions it normally creates
-linkage stubs (PLT call and long branch) using power10 instructions.
-In particular for @code{@@notoc} PLT calls where @code{r2} is not
-known the power10 stubs are smaller and faster, so are preferred for
+linkage stubs (PLT call and long branch) using power10 instructions
+for @code{@@notoc} PLT calls where @code{r2} is not known.  The
+power10 notoc stubs are smaller and faster, so are preferred for
 power10.  @option{--power10-stubs} and @option{--no-power10-stubs}
 allow you to override the linker's selection of stub instructions.
-For example, when linking a shared library that contains cpu-optimized
-versions of functions for both power9 and power10, you might use
-@option{--no-power10-stubs} so that power9 code making calls doesn't
-attempt to execute power10 instructions.
+@option{--power10-stubs=auto} allows the user to select the default
+auto mode.
 @end table
 
 @ifclear GENERIC
diff --git a/ld/testsuite/ld-powerpc/callstub-1.d b/ld/testsuite/ld-powerpc/callstub-1.d
index 21eea76..48efdfb 100644
--- a/ld/testsuite/ld-powerpc/callstub-1.d
+++ b/ld/testsuite/ld-powerpc/callstub-1.d
@@ -1,5 +1,5 @@
 #as: -a64 -mpower10
-#ld: -melf64ppc -shared --plt-align=0 --hash-style=gnu
+#ld: -melf64ppc -shared --plt-align=0 --power10-stubs --hash-style=gnu
 #objdump: -dr -Mpower10
 
 .*
@@ -20,10 +20,10 @@ Disassembly of section \.text:
 .*:	(7d 89 03 a6|a6 03 89 7d) 	mtctr   r12
 .*:	(4e 80 04 20|20 04 80 4e) 	bctr
 #...
-.*:	(4b ff ff c1|c1 ff ff 4b) 	bl      .*\.plt_call\.f1>
+
+.*<_start>:
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f1>
 .*:	(e8 41 00 18|18 00 41 e8) 	ld      r2,24\(r1\)
-.*:	(4b ff ff bd|bd ff ff 4b) 	bl      .*\.plt_call\.f1\+0x4>
-.*:	(4b ff ff cd|cd ff ff 4b) 	bl      .*\.plt_call\.f2>
-.*:	(04 10 00 01|01 00 10 04) 	pld     r3,.*
-.*:	(e4 60 .. ..|.. .. 60 e4) 
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f1\+0x4>
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f2>
 #pass
diff --git a/ld/testsuite/ld-powerpc/callstub-2.d b/ld/testsuite/ld-powerpc/callstub-2.d
index 6036707..5b43783 100644
--- a/ld/testsuite/ld-powerpc/callstub-2.d
+++ b/ld/testsuite/ld-powerpc/callstub-2.d
@@ -27,9 +27,9 @@ Disassembly of section \.text:
 .*:	(7d 89 03 a6|a6 03 89 7d) 	mtctr   r12
 .*:	(4e 80 04 20|20 04 80 4e) 	bctr
 
-#...
-.*:	(4b ff ff bd|bd ff ff 4b) 	bl      .*\.plt_call\.f1>
+.*<_start>:
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f1>
 .*:	(e8 41 00 18|18 00 41 e8) 	ld      r2,24\(r1\)
-.*:	(4b ff ff b9|b9 ff ff 4b) 	bl      .*\.plt_call\.f1\+0x4>
-.*:	(4b ff ff d5|d5 ff ff 4b) 	bl      .*\.plt_call\.f2>
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f1\+0x4>
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f2>
 #pass
diff --git a/ld/testsuite/ld-powerpc/callstub-4.d b/ld/testsuite/ld-powerpc/callstub-4.d
new file mode 100644
index 0000000..12ea0d9
--- /dev/null
+++ b/ld/testsuite/ld-powerpc/callstub-4.d
@@ -0,0 +1,32 @@
+#source: callstub-1.s
+#as: -a64 -mpower10
+#ld: -melf64ppc -shared --plt-align=0 --hash-style=gnu
+#objdump: -dr -Mpower10
+
+.*
+
+Disassembly of section \.text:
+
+.*\.plt_call\.f1>:
+.*:	(04 10 00 01|01 00 10 04) 	pld     r12,.*
+.*:	(e5 80 .. ..|.. .. 80 e5) 
+.*:	(7d 89 03 a6|a6 03 89 7d) 	mtctr   r12
+.*:	(4e 80 04 20|20 04 80 4e) 	bctr
+.*:	(f8 41 00 18|18 00 41 f8) 	std     r2,24\(r1\)
+.*:	(e9 82 80 28|28 80 82 e9) 	ld      r12,-32728\(r2\)
+.*:	(7d 89 03 a6|a6 03 89 7d) 	mtctr   r12
+.*:	(4e 80 04 20|20 04 80 4e) 	bctr
+
+.*\.plt_call\.f2>:
+.*:	(04 10 00 01|01 00 10 04) 	pld     r12,.*
+.*:	(e5 80 .. ..|.. .. 80 e5) 
+.*:	(7d 89 03 a6|a6 03 89 7d) 	mtctr   r12
+.*:	(4e 80 04 20|20 04 80 4e) 	bctr
+#...
+
+.*<_start>:
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f1\+0x10>
+.*:	(e8 41 00 18|18 00 41 e8) 	ld      r2,24\(r1\)
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f1>
+.*:	(4b ff .. ..|.. .. ff 4b) 	bl      .*\.plt_call\.f2>
+#pass
diff --git a/ld/testsuite/ld-powerpc/notoc.d b/ld/testsuite/ld-powerpc/notoc.d
index eaafd01..3555be7 100644
--- a/ld/testsuite/ld-powerpc/notoc.d
+++ b/ld/testsuite/ld-powerpc/notoc.d
@@ -1,6 +1,6 @@
 #source: notoc.s
 #as: -a64
-#ld: --no-plt-localentry -T ext.lnk
+#ld: --no-plt-localentry --no-power10-stubs -T ext.lnk
 #objdump: -d
 #target: powerpc64*-*-*
 
diff --git a/ld/testsuite/ld-powerpc/notoc3.d b/ld/testsuite/ld-powerpc/notoc3.d
new file mode 100644
index 0000000..ce19e99
--- /dev/null
+++ b/ld/testsuite/ld-powerpc/notoc3.d
@@ -0,0 +1,97 @@
+#as: -a64 -mpower10
+#ld: --no-plt-localentry -T ext.lnk
+#objdump: -d
+#target: powerpc64*-*-*
+
+.*
+
+Disassembly of section \.text:
+
+.* <.*\.long_branch\.f1>:
+.*:	(18 00 41 f8|f8 41 00 18) 	std     r2,24\(r1\)
+.*:	(.. .. 00 48|48 00 .. ..) 	b       .* <f1>
+
+.* <.*\.long_branch\.g1>:
+.*:	(18 00 41 f8|f8 41 00 18) 	std     r2,24\(r1\)
+.*:	(.. .. 00 48|48 00 .. ..) 	b       .* <g1>
+
+.* <.*\.plt_branch\.ext>:
+.*:	(00 20 60 3d|3d 60 20 00) 	lis     r11,8192
+.*:	(00 00 6b 61|61 6b 00 00) 	ori     r11,r11,0
+.*:	(ff ef 13 06|06 13 ef ff) 	pla     r12,-268435736
+.*:	(e8 fe 80 39|39 80 fe e8) 
+.*:	(46 17 6b 79|79 6b 17 46) 	rldicr  r11,r11,34,29
+.*:	(14 62 8b 7d|7d 8b 62 14) 	add     r12,r11,r12
+.*:	(a6 03 89 7d|7d 89 03 a6) 	mtctr   r12
+.*:	(20 04 80 4e|4e 80 04 20) 	bctr
+.*:	(00 80 82 e9|e9 82 80 00) 	ld      r12,-32768\(r2\)
+.*:	(a6 03 89 7d|7d 89 03 a6) 	mtctr   r12
+.*:	(20 04 80 4e|4e 80 04 20) 	bctr
+
+.* <.*\.long_branch\.f2>:
+.*:	(00 00 00 60|60 00 00 00) 	nop
+.*:	(00 00 10 06|06 10 00 00) 	pla     r12,108
+.*:	(6c 00 80 39|39 80 00 6c) 
+.*:	(.. .. 00 48|48 00 .. ..) 	b       .* <f2>
+
+.* <.*\.long_branch\.g2>:
+.*:	(00 00 00 60|60 00 00 00) 	nop
+.*:	(00 00 10 06|06 10 00 00) 	pla     r12,144
+.*:	(90 00 80 39|39 80 00 90) 
+.*:	(.. .. 00 48|48 00 .. ..) 	b       .* <g2>
+#...
+
+.* <f1>:
+.*:	(01 00 00 48|48 00 00 01) 	bl      .* <f1>
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.f2>
+.*:	(.. .. 00 48|48 00 .. ..) 	bl      .* <g1>
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.g2>
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.plt_branch\.ext>
+.*:	(20 00 80 4e|4e 80 00 20) 	blr
+
+.* <g1>:
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.f2>
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <f1>
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.g2>
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <g1>
+.*:	(20 00 80 4e|4e 80 00 20) 	blr
+
+.* <f2>:
+.*:	(02 10 40 3c|3c 40 10 02) 	lis     r2,4098
+.*:	(00 90 42 38|38 42 90 00) 	addi    r2,r2,-28672
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.f1>
+.*:	(18 00 41 e8|e8 41 00 18) 	ld      r2,24\(r1\)
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <f2\+0x8>
+.*:	(00 00 00 60|60 00 00 00) 	nop
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.g1>
+.*:	(18 00 41 e8|e8 41 00 18) 	ld      r2,24\(r1\)
+.*:	(.. .. 00 48|48 00 .. ..) 	bl      .* <g2\+0x8>
+.*:	(00 00 00 60|60 00 00 00) 	nop
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.plt_branch\.ext\+0x20>
+.*:	(00 00 00 60|60 00 00 00) 	nop
+.*:	(20 00 80 4e|4e 80 00 20) 	blr
+
+.* <g2>:
+.*:	(02 10 40 3c|3c 40 10 02) 	lis     r2,4098
+.*:	(00 90 42 38|38 42 90 00) 	addi    r2,r2,-28672
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <f2\+0x8>
+.*:	(00 00 00 60|60 00 00 00) 	nop
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.f1>
+.*:	(18 00 41 e8|e8 41 00 18) 	ld      r2,24\(r1\)
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <g2\+0x8>
+.*:	(00 00 00 60|60 00 00 00) 	nop
+.*:	(.. .. ff 4b|4b ff .. ..) 	bl      .* <.*\.long_branch\.g1>
+.*:	(18 00 41 e8|e8 41 00 18) 	ld      r2,24\(r1\)
+.*:	(20 00 80 4e|4e 80 00 20) 	blr
+
+.* <_start>:
+.*:	(00 00 00 48|48 00 00 00) 	b       .* <_start>
+#...
+
+Disassembly of section \.text\.ext:
+
+8000000000000000 <ext>:
+8000000000000000:	(02 10 40 3c|3c 40 10 02) 	lis     r2,4098
+8000000000000004:	(00 90 42 38|38 42 90 00) 	addi    r2,r2,-28672
+8000000000000008:	(00 00 00 60|60 00 00 00) 	nop
+800000000000000c:	(20 00 80 4e|4e 80 00 20) 	blr
diff --git a/ld/testsuite/ld-powerpc/notoc3.s b/ld/testsuite/ld-powerpc/notoc3.s
new file mode 100644
index 0000000..afbefa4
--- /dev/null
+++ b/ld/testsuite/ld-powerpc/notoc3.s
@@ -0,0 +1,59 @@
+ .text
+ .globl f1, f2, g1, g2, _start
+ .weak ext
+ .abiversion 2
+
+f1:
+ .localentry f1,1
+ bl f1@notoc
+ bl f2@notoc
+ bl g1@notoc
+ bl g2@notoc
+ bl ext@notoc
+ blr
+
+g1:
+ .localentry g1,1
+ bl f2@notoc
+ bl f1@notoc
+ bl g2@notoc
+ bl g1@notoc
+ blr
+
+f2:
+0:
+ addis 2,12,.TOC.-0b@ha
+ addi 2,2,.TOC.-0b@l
+ .localentry f2,.-0b
+ bl f1
+ nop
+ bl f2
+ nop
+ bl g1
+ nop
+ bl g2
+ nop
+ bl ext
+ nop
+ blr
+
+g2:
+0:
+ addis 2,12,.TOC.-0b@ha
+ addi 2,2,.TOC.-0b@l
+ .localentry g2,.-0b
+ bl f2
+ nop
+ bl f1
+ nop
+ bl g2
+ nop
+ bl g1
+ nop
+ blr
+
+_start:
+ .cfi_startproc
+ b _start
+ pld 3,_start
+ .cfi_endproc
diff --git a/ld/testsuite/ld-powerpc/notoc3.wf b/ld/testsuite/ld-powerpc/notoc3.wf
new file mode 100644
index 0000000..23904f8
--- /dev/null
+++ b/ld/testsuite/ld-powerpc/notoc3.wf
@@ -0,0 +1,17 @@
+Contents of the \.eh_frame section:
+
+
+00000000 0000000000000010 00000000 CIE
+  Version:               1
+  Augmentation:          "zR"
+  Code alignment factor: 4
+  Data alignment factor: -8
+  Return address column: 65
+  Augmentation data:     1b
+  DW_CFA_def_cfa: r1 ofs 0
+
+00000014 0000000000000010 00000018 FDE cie=00000000 pc=000000001000020c\.\.0000000010000218
+  DW_CFA_nop
+  DW_CFA_nop
+  DW_CFA_nop
+
diff --git a/ld/testsuite/ld-powerpc/powerpc.exp b/ld/testsuite/ld-powerpc/powerpc.exp
index 50553ba..bd269a2 100644
--- a/ld/testsuite/ld-powerpc/powerpc.exp
+++ b/ld/testsuite/ld-powerpc/powerpc.exp
@@ -317,10 +317,14 @@ set ppc64elftests {
     {"ambig shared v1" "-shared -melf64ppc" "" "-a64" {funv1.s} {} "funv1.so"}
     {"ambig shared v2" "-shared -melf64ppc" "" "-a64" {funv2.s} {} "funv2.so"}
     {"notoc ext" "" "" "-a64" {ext.s} {} ""}
-    {"notoc" "-melf64ppc --no-plt-localentry -T ext.lnk" "" "-a64" {notoc.s}
+    {"notoc" "-melf64ppc --no-plt-localentry --no-power10-stubs -T ext.lnk"
+	"" "-a64" {notoc.s}
 	{{objdump -d notoc.d} {readelf {-wf -W} notoc.wf}} "notoc"}
     {"notoc2" "-melf64ppc -shared" "" "-a64 -mpower10" {notoc2.s}
 	{{objdump {-d -Mpower10} notoc2.d}} "notoc2"}
+    {"notoc3" "-melf64ppc --no-plt-localentry -T ext.lnk" ""
+	"-a64 -mpower10" {notoc3.s}
+	{{objdump -d notoc3.d} {readelf {-wf -W} notoc3.wf}} "notoc3"}
     {"pcrelopt" "-melf64ppc --hash-style=gnu" "tmpdir/symtocbase.so"
 	"-a64 -mpower10" {pcrelopt.s}
 	{{objdump {-d -Mpower10} pcrelopt.d}
@@ -403,6 +407,7 @@ if [ supports_ppc64 ] then {
     run_dump_test "callstub-1"
     run_dump_test "callstub-2"
     run_dump_test "callstub-3"
+    run_dump_test "callstub-4"
     run_dump_test "tlsgd"
     run_dump_test "tlsld"
     run_dump_test "tlsie"