Eli Friedman
2013-Jul-10 20:18 UTC
[LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
On Wed, Jul 10, 2013 at 12:29 PM, Ramkumar Ramachandra <artagnon at gmail.com> wrote:> The instructions btr and bts are perfectly valid, and have existed since > Intel 386. GNU as supports them fine. Unfortunately, LLVM does not > support them, and barfs with: > > error: ambiguous instructions require an explicit suffix > > Fix this problem by disambiguating it correctly, following the example > set by 824a907. > > Cc: Eli Friedman <eli.friedman at gmail.com> > Cc: Chris Lattner <sabre at nondot.org> > Signed-off-by: Ramkumar Ramachandra <artagnon at gmail.com> > --- > I've probably done something stupid; seems to build correctly, but > that's all I know. Also, tests are pending. > > lib/Target/X86/X86InstrInfo.td | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/lib/Target/X86/X86InstrInfo.td b/lib/Target/X86/X86InstrInfo.td > index f33ae2a..4950674 100644 > --- a/lib/Target/X86/X86InstrInfo.td > +++ b/lib/Target/X86/X86InstrInfo.td > @@ -1971,6 +1971,10 @@ def : InstAlias<"aam", (AAM8i8 10)>; > // Disambiguate the mem/imm form of bt-without-a-suffix as btl. > def : InstAlias<"bt $imm, $mem", (BT32mi8 i32mem:$mem, i32i8imm:$imm)>; > > +// Disambiguate btr and bts, just like GNU as. > +def : InstAlias<"btr $imm, $mem", (BT16mi8 i16mem:$mem, i16i8imm:$imm)>; > +def : InstAlias<"bts $imm, $mem", (BT16mi8 i16mem:$mem, i16i8imm:$imm)>; > + > // clr aliases. > def : InstAlias<"clrb $reg", (XOR8rr GR8 :$reg, GR8 :$reg)>; > def : InstAlias<"clrw $reg", (XOR16rr GR16:$reg, GR16:$reg)>; > -- > 1.8.3.2.736.g869de25 >Please send patches to llvm-commits. Please include a testcase with each patch. Please check that your patch actually works correctly before sending it to the mailing list for review. (See http://llvm.org/docs/DeveloperPolicy.html .) -Eli
Jim Grosbach
2013-Jul-10 20:41 UTC
[LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
Also, please elaborate on why this is a good change. Because gas accepts it isn’t sufficient reason in and of itself. -Jim On Jul 10, 2013, at 1:18 PM, Eli Friedman <eli.friedman at gmail.com> wrote:> On Wed, Jul 10, 2013 at 12:29 PM, Ramkumar Ramachandra > <artagnon at gmail.com> wrote: >> The instructions btr and bts are perfectly valid, and have existed since >> Intel 386. GNU as supports them fine. Unfortunately, LLVM does not >> support them, and barfs with: >> >> error: ambiguous instructions require an explicit suffix >> >> Fix this problem by disambiguating it correctly, following the example >> set by 824a907. >> >> Cc: Eli Friedman <eli.friedman at gmail.com> >> Cc: Chris Lattner <sabre at nondot.org> >> Signed-off-by: Ramkumar Ramachandra <artagnon at gmail.com> >> --- >> I've probably done something stupid; seems to build correctly, but >> that's all I know. Also, tests are pending. >> >> lib/Target/X86/X86InstrInfo.td | 4 ++++ >> 1 file changed, 4 insertions(+) >> >> diff --git a/lib/Target/X86/X86InstrInfo.td b/lib/Target/X86/X86InstrInfo.td >> index f33ae2a..4950674 100644 >> --- a/lib/Target/X86/X86InstrInfo.td >> +++ b/lib/Target/X86/X86InstrInfo.td >> @@ -1971,6 +1971,10 @@ def : InstAlias<"aam", (AAM8i8 10)>; >> // Disambiguate the mem/imm form of bt-without-a-suffix as btl. >> def : InstAlias<"bt $imm, $mem", (BT32mi8 i32mem:$mem, i32i8imm:$imm)>; >> >> +// Disambiguate btr and bts, just like GNU as. >> +def : InstAlias<"btr $imm, $mem", (BT16mi8 i16mem:$mem, i16i8imm:$imm)>; >> +def : InstAlias<"bts $imm, $mem", (BT16mi8 i16mem:$mem, i16i8imm:$imm)>; >> + >> // clr aliases. >> def : InstAlias<"clrb $reg", (XOR8rr GR8 :$reg, GR8 :$reg)>; >> def : InstAlias<"clrw $reg", (XOR16rr GR16:$reg, GR16:$reg)>; >> -- >> 1.8.3.2.736.g869de25 >> > > Please send patches to llvm-commits. Please include a testcase with > each patch. Please check that your patch actually works correctly > before sending it to the mailing list for review. (See > http://llvm.org/docs/DeveloperPolicy.html .) > > -Eli > _______________________________________________ > LLVM Developers mailing list > LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20130710/c37694d0/attachment.html>
Ramkumar Ramachandra
2013-Jul-10 20:41 UTC
[LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
Eli Friedman wrote:>> I've probably done something stupid; seems to build correctly, but >> that's all I know. Also, tests are pending. > > Please send patches to llvm-commits. Please include a testcase with > each patch. Please check that your patch actually works correctly > before sending it to the mailing list for review.I know. I was asking for a sanity check before doing another iteration. Never mind: I'll submit another iteration soon.
Ramkumar Ramachandra
2013-Jul-10 20:44 UTC
[LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
Jim Grosbach wrote:> Also, please elaborate on why this is a good change. Because gas accepts it > isn’t sufficient reason in and of itself.That they're valid instructions isn't sufficient reason? Should I additionally say that linux.git uses them? I wrote:> The instructions btr and bts are perfectly valid, and have existed since > Intel 386.
Seemingly Similar Threads
- [LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
- [LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
- [LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
- [LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts
- [LLVMdev] [PATCH] x86: disambiguate unqualified btr, bts