ports/multimedia/dav1d/patches/patch-src_x86_itx16_avx512_asm

Index: src/x86/itx16_avx512.asm
--- src/x86/itx16_avx512.asm.orig
+++ src/x86/itx16_avx512.asm
@@ -274,6 +274,7 @@ cglobal inv_txfm_add_%1_%2_%4_10bpc, 4, 7, 0, dst, str
     times ((%%end - %%p1) >> 31) & 1 jmp %%p1
 ALIGN function_align
 %%end:
+    _CET_ENDBR
 %endif
 %endmacro
 
@@ -330,6 +331,7 @@ cglobal idct_8x8_internal_10bpc, 0, 7, 16, dst, stride
     punpckldq            m0, m2
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     lea                  r5, [o_base_8bpc]
     vextracti32x8       ym2, m0, 1
     vextracti32x8       ym3, m1, 1
@@ -461,6 +463,7 @@ cglobal iadst_8x8_internal_10bpc, 0, 7, 16, dst, strid
     vpermt2q             m0, m2, m3
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     call .main_pass2
     movu                m10, [permC+2]
     vbroadcasti32x8     m12, [pw_2048_m2048+16]
@@ -531,6 +534,7 @@ cglobal iflipadst_8x8_internal_10bpc, 0, 7, 16, dst, s
     paddd                m4, m11
     jmp m(iadst_8x8_internal_10bpc).pass1_end
 .pass2:
+    _CET_ENDBR
     call m(iadst_8x8_internal_10bpc).main_pass2
     movu                m10, [permC+1]
     vbroadcasti32x8     m12, [pw_m2048_2048+16]
@@ -561,6 +565,7 @@ cglobal iidentity_8x8_internal_10bpc, 0, 7, 16, dst, s
     punpckhdq            m1, m2        ; 2 3   6 7
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     movu                 m3, [o(permC+2)]
     vpbroadcastd        m12, [o(pw_4096)]
     psrlq                m2, m3, 32
@@ -600,6 +605,7 @@ cglobal idct_8x16_internal_10bpc, 0, 7, 16, dst, strid
     packssdw             m3, m7
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     mova                 m8, [o(idct8x16p)]
     REPX  {vpermb x, m8, x}, m0, m1, m2, m3
     punpckhdq            m5, m0, m1
@@ -829,6 +835,7 @@ cglobal iadst_8x16_internal_10bpc, 0, 7, 16, dst, stri
     vextracti32x8       ym3, m3, 1
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     call .pass2_main
     movu                 m4, [permB+2]
     vbroadcasti32x8     m12, [pw_2048_m2048+16]
@@ -964,6 +971,7 @@ cglobal iflipadst_8x16_internal_10bpc, 0, 7, 16, dst, 
     paddd                m4, m11
     jmp m(iadst_8x16_internal_10bpc).fast_end
 .pass2:
+    _CET_ENDBR
     call m(iadst_8x16_internal_10bpc).pass2_main
     movu                 m7, [permB+2]
     vbroadcasti32x8     m12, [pw_m2048_2048+16]
@@ -985,6 +993,7 @@ cglobal iidentity_8x16_internal_10bpc, 0, 7, 16, dst, 
     call m(idct_8x16_internal_10bpc).load2
     jmp m(idct_8x16_internal_10bpc).pass1_end
 .pass2:
+    _CET_ENDBR
     vpbroadcastd         m8, [o(pw_1697x16)]
     pmulhrsw             m4, m8, m0
     pmulhrsw             m5, m8, m1
@@ -1131,6 +1140,7 @@ cglobal idct_16x8_internal_10bpc, 0, 7, 16, dst, strid
     call .transpose_16x8
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     lea                  r5, [o_base_8bpc]
     call m(idct_16x8_internal_8bpc).main
     movshdup             m4, [permC]
@@ -1270,6 +1280,7 @@ cglobal iadst_16x8_internal_10bpc, 0, 7, 16, dst, stri
     REPX       {psrad x, 1}, m0, m4, m1, m5, m2, m6, m3, m7
     jmp m(idct_16x8_internal_10bpc).pass1_end2
 .pass2:
+    _CET_ENDBR
     call .main_pass2
     vpermq               m8, m11, m0
     vpermq               m9, m11, m1
@@ -1436,6 +1447,7 @@ cglobal iflipadst_16x8_internal_10bpc, 0, 7, 16, dst, 
     psubd                m0, m9, m8
     jmp m(iadst_16x8_internal_10bpc).pass1_end
 .pass2:
+    _CET_ENDBR
     call m(iadst_16x8_internal_10bpc).main_pass2
     psrlq               m11, 8
     vpermq               m8, m11, m3
@@ -1465,6 +1477,7 @@ cglobal iidentity_16x8_internal_10bpc, 0, 7, 16, dst, 
     call m(idct_16x8_internal_10bpc).transpose_16x8
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     movshdup             m4, [o(permC)]
     vpbroadcastd        m11, [o(pw_4096)]
     mova                 m5, m4
@@ -1538,6 +1551,7 @@ cglobal idct_16x16_internal_10bpc, 0, 7, 16, dst, stri
     jge .zero_loop
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     lea                  r5, [o_base_8bpc]
     call m(idct_16x16_internal_8bpc).main
     movshdup            m12, [permC]
@@ -1801,6 +1815,7 @@ cglobal iadst_16x16_internal_10bpc, 0, 7, 16, dst, str
     REPX {mova [cq+64*x], ym4}, 0, 1, 2, 3, 4, 5, 6, 7
     jmp                tx2q
 .pass2:
+    _CET_ENDBR
     lea                  r5, [o_base_8bpc]
     call m(iadst_16x16_internal_8bpc).main_pass2b
     movshdup            m12, [permC]
@@ -2001,6 +2016,7 @@ cglobal iflipadst_16x16_internal_10bpc, 0, 7, 16, dst,
     psubd                m0, m9, m8
     jmp m(iadst_16x16_internal_10bpc).pass1_fast_end
 .pass2:
+    _CET_ENDBR
     lea                  r5, [o_base_8bpc]
     call m(iadst_16x16_internal_8bpc).main_pass2b
     movshdup            m12, [permC]
@@ -2064,6 +2080,7 @@ cglobal iidentity_16x16_internal_10bpc, 0, 7, 16, dst,
     REPX       {mova x, m7}, m4, m5, m6
     jmp m(idct_16x16_internal_10bpc).pass1_end3
 .pass2:
+    _CET_ENDBR
     movshdup            m14, [o(permC)]
     vpbroadcastd        m15, [o(pw_1697x16)]
     lea                  r6, [strideq*3]
@@ -4434,7 +4451,6 @@ cglobal inv_txfm_add_dct_dct_32x64_10bpc, 4, 7, 0, dst
     mova                m10, m13
     call .pass2_fast2_start
 .end:
-
     pxor                m31, m31
 
 .left_zero_loop:
@@ -5050,6 +5066,7 @@ ALIGN function_align
     call .pass2
     RET
 .pass2:
+    _CET_ENDBR
     psrlq               m12, [permC], 24    ;  0  2  8 10  1  3  9 11
     psrlq               m13, m12, 32        ;  4  6 12 14  5  7 13 15
     call m(inv_txfm_add_dct_dct_32x16_10bpc).transpose_16x32