Imported Upstream version 3.2.2

[debian/gnuradio] / gnuradio-core / src / lib / filter / fcomplex_dotprod_sse64.S

diff --git a/gnuradio-core/src/lib/filter/fcomplex_dotprod_sse64.S b/gnuradio-core/src/lib/filter/fcomplex_dotprod_sse64.S
index 64622972a554175798129522db38b44c00b2c69c..9d985a19e8360ccf8e95af16b73d85977c58c7a5 100644 (file)
--- a/gnuradio-core/src/lib/filter/fcomplex_dotprod_sse64.S
+++ b/gnuradio-core/src/lib/filter/fcomplex_dotprod_sse64.S
@@ -5,7 +5,7 @@
 # 
 # GNU Radio is free software; you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 2, or (at your option)
+# the Free Software Foundation; either version 3, or (at your option)
 # any later version.
 # 
 # GNU Radio is distributed in the hope that it will be useful,
@@ -79,7 +79,7 @@ GLOB_SYMB(fcomplex_dotprod_sse):
        jmp     .L1_test
 
        .p2align 4
-.loop1:        
+.Loop1:        
 
        movlps  0(%rdi), %xmm0
        shufps  $0x50, %xmm0, %xmm0     # b01010000
@@ -90,7 +90,7 @@ GLOB_SYMB(fcomplex_dotprod_sse):
        addps   %xmm0, %xmm4
 .L1_test:      
        dec     %rax
-       jge     .loop1
+       jge     .Loop1
 
        
        # set up for primary loop which is unrolled 4 times
@@ -99,7 +99,7 @@ GLOB_SYMB(fcomplex_dotprod_sse):
        movaps  %xmm5, %xmm7 
 
        shr     $2, %rdx                # n_2_complex_blocks / 4
-       je      .cleanup                # if zero, take short path
+       je      .Lcleanup               # if zero, take short path
 
        # finish setup and loop priming
 
@@ -117,7 +117,7 @@ GLOB_SYMB(fcomplex_dotprod_sse):
        # hence enter loop at top
 
        .p2align 4
-.loop2:
+.Loop2:
        addps   %xmm2, %xmm6
        movlps  0x10(%rdi), %xmm2
 
@@ -149,7 +149,7 @@ GLOB_SYMB(fcomplex_dotprod_sse):
        add     $0x40, %rsi
        add     $0x20, %rdi
        dec     %rdx
-       jne     .loop2
+       jne     .Loop2
 
        # OK, now we've done with all the multiplies, but
        # we still need to handle the unaccumulated
@@ -168,7 +168,7 @@ GLOB_SYMB(fcomplex_dotprod_sse):
        # At this point, xmm4 contains 2x2 partial sums.  We need
        # to compute a "horizontal complex add" across xmm4.  
        
-.cleanup:                              # xmm4 = r1 i2 r3 i4
+.Lcleanup:                             # xmm4 = r1 i2 r3 i4
        movhlps %xmm4, %xmm0            # xmm0 = ?? ?? r1 r2
        addps   %xmm4, %xmm0            # xmm0 = ?? ?? r1+r3 i2+i4
        movlps  %xmm0, (%rcx)           # store low 2x32 bits (complex) to memory
@@ -177,3 +177,7 @@ GLOB_SYMB(fcomplex_dotprod_sse):
 
 FUNC_TAIL(fcomplex_dotprod_sse)
        .ident  "Hand coded x86_64 SSE assembly"
+
+#if defined(__linux__) && defined(__ELF__)
+.section .note.GNU-stack,"",%progbits
+#endif