Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux
1
fork

Configure Feed

Select the types of activity you want to include in your feed.

drm/amdkfd: Fix instruction hazard in gfx12 trap handler

VALU instructions with SGPR source need wait states to avoid hazard
with SALU using different SGPR.

v2: Eliminate some hazards to reduce code explosion

Signed-off-by: Jay Cornwall <jay.cornwall@amd.com>
Reviewed-by: Lancelot Six <lancelot.six@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

authored by

Jay Cornwall and committed by
Alex Deucher
7e0459d4 18537feb

+417 -368
+373 -330
drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
··· 3644 3644 }; 3645 3645 3646 3646 static const uint32_t cwsr_trap_gfx12_hex[] = { 3647 - 0xbfa00001, 0xbfa0024b, 3647 + 0xbfa00001, 0xbfa002a2, 3648 3648 0xb0804009, 0xb8f8f804, 3649 3649 0x9178ff78, 0x00008c00, 3650 3650 0xb8fbf811, 0x8b6eff78, ··· 3718 3718 0x00011677, 0xd7610000, 3719 3719 0x00011a79, 0xd7610000, 3720 3720 0x00011c7e, 0xd7610000, 3721 - 0x00011e7f, 0xbefe00ff, 3721 + 0x00011e7f, 0xd8500000, 3722 + 0x00000000, 0xd8500000, 3723 + 0x00000000, 0xd8500000, 3724 + 0x00000000, 0xd8500000, 3725 + 0x00000000, 0xd8500000, 3726 + 0x00000000, 0xd8500000, 3727 + 0x00000000, 0xd8500000, 3728 + 0x00000000, 0xd8500000, 3729 + 0x00000000, 0xbefe00ff, 3722 3730 0x00003fff, 0xbeff0080, 3723 3731 0xee0a407a, 0x000c0000, 3724 3732 0x00004000, 0xd760007a, ··· 3763 3755 0x00000200, 0xbef600ff, 3764 3756 0x01000000, 0x7e000280, 3765 3757 0x7e020280, 0x7e040280, 3766 - 0xbefd0080, 0xbe804ec2, 3767 - 0xbf94fffe, 0xb8faf804, 3768 - 0x8b7a847a, 0x91788478, 3769 - 0x8c787a78, 0xd7610002, 3770 - 0x0000fa71, 0x807d817d, 3771 - 0xd7610002, 0x0000fa6c, 3772 - 0x807d817d, 0x917aff6d, 3773 - 0x80000000, 0xd7610002, 3774 - 0x0000fa7a, 0x807d817d, 3775 - 0xd7610002, 0x0000fa6e, 3776 - 0x807d817d, 0xd7610002, 3777 - 0x0000fa6f, 0x807d817d, 3778 - 0xd7610002, 0x0000fa78, 3779 - 0x807d817d, 0xb8faf811, 3780 - 0xd7610002, 0x0000fa7a, 3781 - 0x807d817d, 0xd7610002, 3782 - 0x0000fa7b, 0x807d817d, 3783 - 0xb8f1f801, 0xd7610002, 3784 - 0x0000fa71, 0x807d817d, 3785 - 0xb8f1f814, 0xd7610002, 3786 - 0x0000fa71, 0x807d817d, 3787 - 0xb8f1f815, 0xd7610002, 3788 - 0x0000fa71, 0x807d817d, 3789 - 0xb8f1f812, 0xd7610002, 3790 - 0x0000fa71, 0x807d817d, 3791 - 0xb8f1f813, 0xd7610002, 3792 - 0x0000fa71, 0x807d817d, 3758 + 0xbe804ec2, 0xbf94fffe, 3759 + 0xb8faf804, 0x8b7a847a, 3760 + 0x91788478, 0x8c787a78, 3761 + 0x917aff6d, 0x80000000, 3762 + 0xd7610002, 0x00010071, 3763 + 0xd7610002, 0x0001026c, 3764 + 0xd7610002, 0x0001047a, 3765 + 0xd7610002, 0x0001066e, 3766 + 0xd7610002, 0x0001086f, 3767 + 0xd7610002, 0x00010a78, 3768 + 0xd7610002, 0x00010e7b, 3769 + 0xd8500000, 0x00000000, 3770 + 0xd8500000, 0x00000000, 3771 + 0xd8500000, 0x00000000, 3772 + 0xd8500000, 0x00000000, 3773 + 0xd8500000, 0x00000000, 3774 + 0xd8500000, 0x00000000, 3775 + 0xd8500000, 0x00000000, 3776 + 0xd8500000, 0x00000000, 3777 + 0xb8faf811, 0xd7610002, 3778 + 0x00010c7a, 0xb8faf801, 3779 + 0xd7610002, 0x0001107a, 3780 + 0xb8faf814, 0xd7610002, 3781 + 0x0001127a, 0xb8faf815, 3782 + 0xd7610002, 0x0001147a, 3783 + 0xb8faf812, 0xd7610002, 3784 + 0x0001167a, 0xb8faf813, 3785 + 0xd7610002, 0x0001187a, 3793 3786 0xb8faf802, 0xd7610002, 3794 - 0x0000fa7a, 0x807d817d, 3795 - 0xbefa50c1, 0xbfc70000, 3796 - 0xd7610002, 0x0000fa7a, 3797 - 0x807d817d, 0xbefe00ff, 3787 + 0x00011a7a, 0xbefa50c1, 3788 + 0xbfc70000, 0xd7610002, 3789 + 0x00011c7a, 0xd8500000, 3790 + 0x00000000, 0xd8500000, 3791 + 0x00000000, 0xd8500000, 3792 + 0x00000000, 0xd8500000, 3793 + 0x00000000, 0xd8500000, 3794 + 0x00000000, 0xd8500000, 3795 + 0x00000000, 0xd8500000, 3796 + 0x00000000, 0xd8500000, 3797 + 0x00000000, 0xbefe00ff, 3798 3798 0x0000ffff, 0xbeff0080, 3799 3799 0xc4068070, 0x008ce802, 3800 3800 0x00000000, 0xbefe00c1, ··· 3817 3801 0xbe824102, 0xbe844104, 3818 3802 0xbe864106, 0xbe884108, 3819 3803 0xbe8a410a, 0xbe8c410c, 3820 - 0xbe8e410e, 0xd7610002, 3821 - 0x0000f200, 0x80798179, 3822 - 0xd7610002, 0x0000f201, 3823 - 0x80798179, 0xd7610002, 3824 - 0x0000f202, 0x80798179, 3825 - 0xd7610002, 0x0000f203, 3826 - 0x80798179, 0xd7610002, 3827 - 0x0000f204, 0x80798179, 3828 - 0xd7610002, 0x0000f205, 3829 - 0x80798179, 0xd7610002, 3830 - 0x0000f206, 0x80798179, 3831 - 0xd7610002, 0x0000f207, 3832 - 0x80798179, 0xd7610002, 3833 - 0x0000f208, 0x80798179, 3834 - 0xd7610002, 0x0000f209, 3835 - 0x80798179, 0xd7610002, 3836 - 0x0000f20a, 0x80798179, 3837 - 0xd7610002, 0x0000f20b, 3838 - 0x80798179, 0xd7610002, 3839 - 0x0000f20c, 0x80798179, 3840 - 0xd7610002, 0x0000f20d, 3841 - 0x80798179, 0xd7610002, 3842 - 0x0000f20e, 0x80798179, 3843 - 0xd7610002, 0x0000f20f, 3844 - 0x80798179, 0xbf06a079, 3845 - 0xbfa10007, 0xc4068070, 3846 - 0x008ce802, 0x00000000, 3847 - 0x8070ff70, 0x00000080, 3848 - 0xbef90080, 0x7e040280, 3849 - 0x807d907d, 0xbf0aff7d, 3850 - 0x00000060, 0xbfa2ffbb, 3851 - 0xbe804100, 0xbe824102, 3852 - 0xbe844104, 0xbe864106, 3853 - 0xbe884108, 0xbe8a410a, 3854 - 0xd7610002, 0x0000f200, 3855 - 0x80798179, 0xd7610002, 3856 - 0x0000f201, 0x80798179, 3857 - 0xd7610002, 0x0000f202, 3858 - 0x80798179, 0xd7610002, 3859 - 0x0000f203, 0x80798179, 3860 - 0xd7610002, 0x0000f204, 3861 - 0x80798179, 0xd7610002, 3862 - 0x0000f205, 0x80798179, 3863 - 0xd7610002, 0x0000f206, 3864 - 0x80798179, 0xd7610002, 3865 - 0x0000f207, 0x80798179, 3866 - 0xd7610002, 0x0000f208, 3867 - 0x80798179, 0xd7610002, 3868 - 0x0000f209, 0x80798179, 3869 - 0xd7610002, 0x0000f20a, 3870 - 0x80798179, 0xd7610002, 3871 - 0x0000f20b, 0x80798179, 3804 + 0xbe8e410e, 0xbf068079, 3805 + 0xbfa10032, 0xd7610002, 3806 + 0x00010000, 0xd7610002, 3807 + 0x00010201, 0xd7610002, 3808 + 0x00010402, 0xd7610002, 3809 + 0x00010603, 0xd7610002, 3810 + 0x00010804, 0xd7610002, 3811 + 0x00010a05, 0xd7610002, 3812 + 0x00010c06, 0xd7610002, 3813 + 0x00010e07, 0xd7610002, 3814 + 0x00011008, 0xd7610002, 3815 + 0x00011209, 0xd7610002, 3816 + 0x0001140a, 0xd7610002, 3817 + 0x0001160b, 0xd7610002, 3818 + 0x0001180c, 0xd7610002, 3819 + 0x00011a0d, 0xd7610002, 3820 + 0x00011c0e, 0xd7610002, 3821 + 0x00011e0f, 0xd8500000, 3822 + 0x00000000, 0xd8500000, 3823 + 0x00000000, 0xd8500000, 3824 + 0x00000000, 0xd8500000, 3825 + 0x00000000, 0xd8500000, 3826 + 0x00000000, 0xd8500000, 3827 + 0x00000000, 0xd8500000, 3828 + 0x00000000, 0xd8500000, 3829 + 0x00000000, 0x80799079, 3830 + 0xbfa00038, 0xd7610002, 3831 + 0x00012000, 0xd7610002, 3832 + 0x00012201, 0xd7610002, 3833 + 0x00012402, 0xd7610002, 3834 + 0x00012603, 0xd7610002, 3835 + 0x00012804, 0xd7610002, 3836 + 0x00012a05, 0xd7610002, 3837 + 0x00012c06, 0xd7610002, 3838 + 0x00012e07, 0xd7610002, 3839 + 0x00013008, 0xd7610002, 3840 + 0x00013209, 0xd7610002, 3841 + 0x0001340a, 0xd7610002, 3842 + 0x0001360b, 0xd7610002, 3843 + 0x0001380c, 0xd7610002, 3844 + 0x00013a0d, 0xd7610002, 3845 + 0x00013c0e, 0xd7610002, 3846 + 0x00013e0f, 0xd8500000, 3847 + 0x00000000, 0xd8500000, 3848 + 0x00000000, 0xd8500000, 3849 + 0x00000000, 0xd8500000, 3850 + 0x00000000, 0xd8500000, 3851 + 0x00000000, 0xd8500000, 3852 + 0x00000000, 0xd8500000, 3853 + 0x00000000, 0xd8500000, 3854 + 0x00000000, 0x80799079, 3872 3855 0xc4068070, 0x008ce802, 3873 - 0x00000000, 0xbefe00c1, 3874 - 0x857d9973, 0x8b7d817d, 3875 - 0xbf06817d, 0xbfa20002, 3876 - 0xbeff0080, 0xbfa00001, 3877 - 0xbeff00c1, 0xb8fb4306, 3878 - 0x8b7bc17b, 0xbfa10044, 3879 - 0x8b7aff6d, 0x80000000, 3880 - 0xbfa10041, 0x847b897b, 3881 - 0xbef6007b, 0xb8f03b05, 3882 - 0x80708170, 0xbf0d9973, 3883 - 0xbfa20002, 0x84708970, 3884 - 0xbfa00001, 0x84708a70, 3885 - 0xb8fa1e06, 0x847a8a7a, 3886 - 0x80707a70, 0x8070ff70, 3887 - 0x00000200, 0x8070ff70, 3888 - 0x00000080, 0xbef600ff, 3889 - 0x01000000, 0xd71f0000, 3890 - 0x000100c1, 0xd7200000, 3891 - 0x000200c1, 0x16000084, 3892 - 0x857d9973, 0x8b7d817d, 3893 - 0xbf06817d, 0xbefd0080, 3894 - 0xbfa20013, 0xbe8300ff, 3895 - 0x00000080, 0xbf800000, 3896 - 0xbf800000, 0xbf800000, 3897 - 0xd8d80000, 0x01000000, 3898 - 0xbf8a0000, 0xc4068070, 3899 - 0x008ce801, 0x00000000, 3900 - 0x807d037d, 0x80700370, 3901 - 0xd5250000, 0x0001ff00, 3902 - 0x00000080, 0xbf0a7b7d, 3903 - 0xbfa2fff3, 0xbfa00012, 3904 - 0xbe8300ff, 0x00000100, 3856 + 0x00000000, 0x8070ff70, 3857 + 0x00000080, 0xbef90080, 3858 + 0x7e040280, 0x807d907d, 3859 + 0xbf0aff7d, 0x00000060, 3860 + 0xbfa2ff88, 0xbe804100, 3861 + 0xbe824102, 0xbe844104, 3862 + 0xbe864106, 0xbe884108, 3863 + 0xbe8a410a, 0xd7610002, 3864 + 0x00010000, 0xd7610002, 3865 + 0x00010201, 0xd7610002, 3866 + 0x00010402, 0xd7610002, 3867 + 0x00010603, 0xd7610002, 3868 + 0x00010804, 0xd7610002, 3869 + 0x00010a05, 0xd7610002, 3870 + 0x00010c06, 0xd7610002, 3871 + 0x00010e07, 0xd7610002, 3872 + 0x00011008, 0xd7610002, 3873 + 0x00011209, 0xd7610002, 3874 + 0x0001140a, 0xd7610002, 3875 + 0x0001160b, 0xd8500000, 3876 + 0x00000000, 0xd8500000, 3877 + 0x00000000, 0xd8500000, 3878 + 0x00000000, 0xd8500000, 3879 + 0x00000000, 0xd8500000, 3880 + 0x00000000, 0xd8500000, 3881 + 0x00000000, 0xd8500000, 3882 + 0x00000000, 0xd8500000, 3883 + 0x00000000, 0xc4068070, 3884 + 0x008ce802, 0x00000000, 3885 + 0xbefe00c1, 0x857d9973, 3886 + 0x8b7d817d, 0xbf06817d, 3887 + 0xbfa20002, 0xbeff0080, 3888 + 0xbfa00001, 0xbeff00c1, 3889 + 0xb8fb4306, 0x8b7bc17b, 3890 + 0xbfa10044, 0x8b7aff6d, 3891 + 0x80000000, 0xbfa10041, 3892 + 0x847b897b, 0xbef6007b, 3893 + 0xb8f03b05, 0x80708170, 3894 + 0xbf0d9973, 0xbfa20002, 3895 + 0x84708970, 0xbfa00001, 3896 + 0x84708a70, 0xb8fa1e06, 3897 + 0x847a8a7a, 0x80707a70, 3898 + 0x8070ff70, 0x00000200, 3899 + 0x8070ff70, 0x00000080, 3900 + 0xbef600ff, 0x01000000, 3901 + 0xd71f0000, 0x000100c1, 3902 + 0xd7200000, 0x000200c1, 3903 + 0x16000084, 0x857d9973, 3904 + 0x8b7d817d, 0xbf06817d, 3905 + 0xbefd0080, 0xbfa20013, 3906 + 0xbe8300ff, 0x00000080, 3905 3907 0xbf800000, 0xbf800000, 3906 3908 0xbf800000, 0xd8d80000, 3907 3909 0x01000000, 0xbf8a0000, 3908 3910 0xc4068070, 0x008ce801, 3909 3911 0x00000000, 0x807d037d, 3910 3912 0x80700370, 0xd5250000, 3911 - 0x0001ff00, 0x00000100, 3913 + 0x0001ff00, 0x00000080, 3912 3914 0xbf0a7b7d, 0xbfa2fff3, 3913 - 0xbefe00c1, 0x857d9973, 3914 - 0x8b7d817d, 0xbf06817d, 3915 - 0xbfa20004, 0xbef000ff, 3916 - 0x00000200, 0xbeff0080, 3917 - 0xbfa00003, 0xbef000ff, 3918 - 0x00000400, 0xbeff00c1, 3919 - 0xb8fb3b05, 0x807b817b, 3920 - 0x847b827b, 0x857d9973, 3921 - 0x8b7d817d, 0xbf06817d, 3922 - 0xbfa2001b, 0xbef600ff, 3923 - 0x01000000, 0xbefd0084, 3924 - 0xbf0a7b7d, 0xbfa10040, 3925 - 0x7e008700, 0x7e028701, 3926 - 0x7e048702, 0x7e068703, 3927 - 0xc4068070, 0x008ce800, 3928 - 0x00000000, 0xc4068070, 3929 - 0x008ce801, 0x00008000, 3930 - 0xc4068070, 0x008ce802, 3931 - 0x00010000, 0xc4068070, 3932 - 0x008ce803, 0x00018000, 3933 - 0x807d847d, 0x8070ff70, 3934 - 0x00000200, 0xbf0a7b7d, 3935 - 0xbfa2ffeb, 0xbfa0002a, 3915 + 0xbfa00012, 0xbe8300ff, 3916 + 0x00000100, 0xbf800000, 3917 + 0xbf800000, 0xbf800000, 3918 + 0xd8d80000, 0x01000000, 3919 + 0xbf8a0000, 0xc4068070, 3920 + 0x008ce801, 0x00000000, 3921 + 0x807d037d, 0x80700370, 3922 + 0xd5250000, 0x0001ff00, 3923 + 0x00000100, 0xbf0a7b7d, 3924 + 0xbfa2fff3, 0xbefe00c1, 3925 + 0x857d9973, 0x8b7d817d, 3926 + 0xbf06817d, 0xbfa20004, 3927 + 0xbef000ff, 0x00000200, 3928 + 0xbeff0080, 0xbfa00003, 3929 + 0xbef000ff, 0x00000400, 3930 + 0xbeff00c1, 0xb8fb3b05, 3931 + 0x807b817b, 0x847b827b, 3932 + 0x857d9973, 0x8b7d817d, 3933 + 0xbf06817d, 0xbfa2001b, 3936 3934 0xbef600ff, 0x01000000, 3937 3935 0xbefd0084, 0xbf0a7b7d, 3938 - 0xbfa10015, 0x7e008700, 3936 + 0xbfa10040, 0x7e008700, 3939 3937 0x7e028701, 0x7e048702, 3940 3938 0x7e068703, 0xc4068070, 3941 3939 0x008ce800, 0x00000000, 3942 3940 0xc4068070, 0x008ce801, 3943 - 0x00010000, 0xc4068070, 3944 - 0x008ce802, 0x00020000, 3941 + 0x00008000, 0xc4068070, 3942 + 0x008ce802, 0x00010000, 3945 3943 0xc4068070, 0x008ce803, 3946 - 0x00030000, 0x807d847d, 3947 - 0x8070ff70, 0x00000400, 3944 + 0x00018000, 0x807d847d, 3945 + 0x8070ff70, 0x00000200, 3948 3946 0xbf0a7b7d, 0xbfa2ffeb, 3949 - 0xb8fb1e06, 0x8b7bc17b, 3950 - 0xbfa1000d, 0x847b837b, 3951 - 0x807b7d7b, 0xbefe00c1, 3952 - 0xbeff0080, 0x7e008700, 3947 + 0xbfa0002a, 0xbef600ff, 3948 + 0x01000000, 0xbefd0084, 3949 + 0xbf0a7b7d, 0xbfa10015, 3950 + 0x7e008700, 0x7e028701, 3951 + 0x7e048702, 0x7e068703, 3953 3952 0xc4068070, 0x008ce800, 3954 - 0x00000000, 0x807d817d, 3955 - 0x8070ff70, 0x00000080, 3956 - 0xbf0a7b7d, 0xbfa2fff7, 3957 - 0xbfa0016e, 0xbef4007e, 3958 - 0x8b75ff7f, 0x0000ffff, 3959 - 0x8c75ff75, 0x00040000, 3960 - 0xbef60080, 0xbef700ff, 3961 - 0x10807fac, 0xbef1007f, 3962 - 0xb8f20742, 0x84729972, 3963 - 0x8b6eff7f, 0x04000000, 3964 - 0xbfa1003b, 0xbefe00c1, 3965 - 0x857d9972, 0x8b7d817d, 3966 - 0xbf06817d, 0xbfa20002, 3967 - 0xbeff0080, 0xbfa00001, 3968 - 0xbeff00c1, 0xb8ef4306, 3969 - 0x8b6fc16f, 0xbfa10030, 3970 - 0x846f896f, 0xbef6006f, 3953 + 0x00000000, 0xc4068070, 3954 + 0x008ce801, 0x00010000, 3955 + 0xc4068070, 0x008ce802, 3956 + 0x00020000, 0xc4068070, 3957 + 0x008ce803, 0x00030000, 3958 + 0x807d847d, 0x8070ff70, 3959 + 0x00000400, 0xbf0a7b7d, 3960 + 0xbfa2ffeb, 0xb8fb1e06, 3961 + 0x8b7bc17b, 0xbfa1000d, 3962 + 0x847b837b, 0x807b7d7b, 3963 + 0xbefe00c1, 0xbeff0080, 3964 + 0x7e008700, 0xc4068070, 3965 + 0x008ce800, 0x00000000, 3966 + 0x807d817d, 0x8070ff70, 3967 + 0x00000080, 0xbf0a7b7d, 3968 + 0xbfa2fff7, 0xbfa0016e, 3969 + 0xbef4007e, 0x8b75ff7f, 3970 + 0x0000ffff, 0x8c75ff75, 3971 + 0x00040000, 0xbef60080, 3972 + 0xbef700ff, 0x10807fac, 3973 + 0xbef1007f, 0xb8f20742, 3974 + 0x84729972, 0x8b6eff7f, 3975 + 0x04000000, 0xbfa1003b, 3976 + 0xbefe00c1, 0x857d9972, 3977 + 0x8b7d817d, 0xbf06817d, 3978 + 0xbfa20002, 0xbeff0080, 3979 + 0xbfa00001, 0xbeff00c1, 3980 + 0xb8ef4306, 0x8b6fc16f, 3981 + 0xbfa10030, 0x846f896f, 3982 + 0xbef6006f, 0xb8f83b05, 3983 + 0x80788178, 0xbf0d9972, 3984 + 0xbfa20002, 0x84788978, 3985 + 0xbfa00001, 0x84788a78, 3986 + 0xb8ee1e06, 0x846e8a6e, 3987 + 0x80786e78, 0x8078ff78, 3988 + 0x00000200, 0x8078ff78, 3989 + 0x00000080, 0xbef600ff, 3990 + 0x01000000, 0x857d9972, 3991 + 0x8b7d817d, 0xbf06817d, 3992 + 0xbefd0080, 0xbfa2000d, 3993 + 0xc4050078, 0x0080e800, 3994 + 0x00000000, 0xbf8a0000, 3995 + 0xdac00000, 0x00000000, 3996 + 0x807dff7d, 0x00000080, 3997 + 0x8078ff78, 0x00000080, 3998 + 0xbf0a6f7d, 0xbfa2fff4, 3999 + 0xbfa0000c, 0xc4050078, 4000 + 0x0080e800, 0x00000000, 4001 + 0xbf8a0000, 0xdac00000, 4002 + 0x00000000, 0x807dff7d, 4003 + 0x00000100, 0x8078ff78, 4004 + 0x00000100, 0xbf0a6f7d, 4005 + 0xbfa2fff4, 0xbef80080, 4006 + 0xbefe00c1, 0x857d9972, 4007 + 0x8b7d817d, 0xbf06817d, 4008 + 0xbfa20002, 0xbeff0080, 4009 + 0xbfa00001, 0xbeff00c1, 4010 + 0xb8ef3b05, 0x806f816f, 4011 + 0x846f826f, 0x857d9972, 4012 + 0x8b7d817d, 0xbf06817d, 4013 + 0xbfa2002c, 0xbef600ff, 4014 + 0x01000000, 0xbeee0078, 4015 + 0x8078ff78, 0x00000200, 4016 + 0xbefd0084, 0xbf0a6f7d, 4017 + 0xbfa10061, 0xc4050078, 4018 + 0x008ce800, 0x00000000, 4019 + 0xc4050078, 0x008ce801, 4020 + 0x00008000, 0xc4050078, 4021 + 0x008ce802, 0x00010000, 4022 + 0xc4050078, 0x008ce803, 4023 + 0x00018000, 0xbf8a0000, 4024 + 0x7e008500, 0x7e028501, 4025 + 0x7e048502, 0x7e068503, 4026 + 0x807d847d, 0x8078ff78, 4027 + 0x00000200, 0xbf0a6f7d, 4028 + 0xbfa2ffea, 0xc405006e, 4029 + 0x008ce800, 0x00000000, 4030 + 0xc405006e, 0x008ce801, 4031 + 0x00008000, 0xc405006e, 4032 + 0x008ce802, 0x00010000, 4033 + 0xc405006e, 0x008ce803, 4034 + 0x00018000, 0xbf8a0000, 4035 + 0xbfa0003d, 0xbef600ff, 4036 + 0x01000000, 0xbeee0078, 4037 + 0x8078ff78, 0x00000400, 4038 + 0xbefd0084, 0xbf0a6f7d, 4039 + 0xbfa10016, 0xc4050078, 4040 + 0x008ce800, 0x00000000, 4041 + 0xc4050078, 0x008ce801, 4042 + 0x00010000, 0xc4050078, 4043 + 0x008ce802, 0x00020000, 4044 + 0xc4050078, 0x008ce803, 4045 + 0x00030000, 0xbf8a0000, 4046 + 0x7e008500, 0x7e028501, 4047 + 0x7e048502, 0x7e068503, 4048 + 0x807d847d, 0x8078ff78, 4049 + 0x00000400, 0xbf0a6f7d, 4050 + 0xbfa2ffea, 0xb8ef1e06, 4051 + 0x8b6fc16f, 0xbfa1000f, 4052 + 0x846f836f, 0x806f7d6f, 4053 + 0xbefe00c1, 0xbeff0080, 4054 + 0xc4050078, 0x008ce800, 4055 + 0x00000000, 0xbf8a0000, 4056 + 0x7e008500, 0x807d817d, 4057 + 0x8078ff78, 0x00000080, 4058 + 0xbf0a6f7d, 0xbfa2fff6, 4059 + 0xbeff00c1, 0xc405006e, 4060 + 0x008ce800, 0x00000000, 4061 + 0xc405006e, 0x008ce801, 4062 + 0x00010000, 0xc405006e, 4063 + 0x008ce802, 0x00020000, 4064 + 0xc405006e, 0x008ce803, 4065 + 0x00030000, 0xbf8a0000, 3971 4066 0xb8f83b05, 0x80788178, 3972 4067 0xbf0d9972, 0xbfa20002, 3973 4068 0x84788978, 0xbfa00001, 3974 4069 0x84788a78, 0xb8ee1e06, 3975 4070 0x846e8a6e, 0x80786e78, 3976 4071 0x8078ff78, 0x00000200, 3977 - 0x8078ff78, 0x00000080, 4072 + 0x80f8ff78, 0x00000050, 3978 4073 0xbef600ff, 0x01000000, 3979 - 0x857d9972, 0x8b7d817d, 3980 - 0xbf06817d, 0xbefd0080, 3981 - 0xbfa2000d, 0xc4050078, 3982 - 0x0080e800, 0x00000000, 3983 - 0xbf8a0000, 0xdac00000, 3984 - 0x00000000, 0x807dff7d, 3985 - 0x00000080, 0x8078ff78, 3986 - 0x00000080, 0xbf0a6f7d, 3987 - 0xbfa2fff4, 0xbfa0000c, 3988 - 0xc4050078, 0x0080e800, 3989 - 0x00000000, 0xbf8a0000, 3990 - 0xdac00000, 0x00000000, 3991 - 0x807dff7d, 0x00000100, 3992 - 0x8078ff78, 0x00000100, 3993 - 0xbf0a6f7d, 0xbfa2fff4, 3994 - 0xbef80080, 0xbefe00c1, 3995 - 0x857d9972, 0x8b7d817d, 3996 - 0xbf06817d, 0xbfa20002, 3997 - 0xbeff0080, 0xbfa00001, 3998 - 0xbeff00c1, 0xb8ef3b05, 3999 - 0x806f816f, 0x846f826f, 4000 - 0x857d9972, 0x8b7d817d, 4001 - 0xbf06817d, 0xbfa2002c, 4002 - 0xbef600ff, 0x01000000, 4003 - 0xbeee0078, 0x8078ff78, 4004 - 0x00000200, 0xbefd0084, 4005 - 0xbf0a6f7d, 0xbfa10061, 4006 - 0xc4050078, 0x008ce800, 4007 - 0x00000000, 0xc4050078, 4008 - 0x008ce801, 0x00008000, 4009 - 0xc4050078, 0x008ce802, 4010 - 0x00010000, 0xc4050078, 4011 - 0x008ce803, 0x00018000, 4012 - 0xbf8a0000, 0x7e008500, 4013 - 0x7e028501, 0x7e048502, 4014 - 0x7e068503, 0x807d847d, 4015 - 0x8078ff78, 0x00000200, 4016 - 0xbf0a6f7d, 0xbfa2ffea, 4017 - 0xc405006e, 0x008ce800, 4018 - 0x00000000, 0xc405006e, 4019 - 0x008ce801, 0x00008000, 4020 - 0xc405006e, 0x008ce802, 4021 - 0x00010000, 0xc405006e, 4022 - 0x008ce803, 0x00018000, 4023 - 0xbf8a0000, 0xbfa0003d, 4024 - 0xbef600ff, 0x01000000, 4025 - 0xbeee0078, 0x8078ff78, 4026 - 0x00000400, 0xbefd0084, 4027 - 0xbf0a6f7d, 0xbfa10016, 4028 - 0xc4050078, 0x008ce800, 4029 - 0x00000000, 0xc4050078, 4030 - 0x008ce801, 0x00010000, 4031 - 0xc4050078, 0x008ce802, 4032 - 0x00020000, 0xc4050078, 4033 - 0x008ce803, 0x00030000, 4034 - 0xbf8a0000, 0x7e008500, 4035 - 0x7e028501, 0x7e048502, 4036 - 0x7e068503, 0x807d847d, 4037 - 0x8078ff78, 0x00000400, 4038 - 0xbf0a6f7d, 0xbfa2ffea, 4039 - 0xb8ef1e06, 0x8b6fc16f, 4040 - 0xbfa1000f, 0x846f836f, 4041 - 0x806f7d6f, 0xbefe00c1, 4042 - 0xbeff0080, 0xc4050078, 4043 - 0x008ce800, 0x00000000, 4044 - 0xbf8a0000, 0x7e008500, 4045 - 0x807d817d, 0x8078ff78, 4046 - 0x00000080, 0xbf0a6f7d, 4047 - 0xbfa2fff6, 0xbeff00c1, 4048 - 0xc405006e, 0x008ce800, 4049 - 0x00000000, 0xc405006e, 4050 - 0x008ce801, 0x00010000, 4051 - 0xc405006e, 0x008ce802, 4052 - 0x00020000, 0xc405006e, 4053 - 0x008ce803, 0x00030000, 4054 - 0xbf8a0000, 0xb8f83b05, 4055 - 0x80788178, 0xbf0d9972, 4056 - 0xbfa20002, 0x84788978, 4057 - 0xbfa00001, 0x84788a78, 4058 - 0xb8ee1e06, 0x846e8a6e, 4059 - 0x80786e78, 0x8078ff78, 4060 - 0x00000200, 0x80f8ff78, 4061 - 0x00000050, 0xbef600ff, 4062 - 0x01000000, 0xbefd00ff, 4063 - 0x0000006c, 0x80f89078, 4064 - 0xf462403a, 0xf0000000, 4065 - 0xbf8a0000, 0x80fd847d, 4066 - 0xbf800000, 0xbe804300, 4067 - 0xbe824302, 0x80f8a078, 4068 - 0xf462603a, 0xf0000000, 4069 - 0xbf8a0000, 0x80fd887d, 4070 - 0xbf800000, 0xbe804300, 4071 - 0xbe824302, 0xbe844304, 4072 - 0xbe864306, 0x80f8c078, 4073 - 0xf462803a, 0xf0000000, 4074 - 0xbf8a0000, 0x80fd907d, 4075 - 0xbf800000, 0xbe804300, 4076 - 0xbe824302, 0xbe844304, 4077 - 0xbe864306, 0xbe884308, 4078 - 0xbe8a430a, 0xbe8c430c, 4079 - 0xbe8e430e, 0xbf06807d, 4080 - 0xbfa1fff0, 0xb980f801, 4081 - 0x00000000, 0xb8f83b05, 4082 - 0x80788178, 0xbf0d9972, 4083 - 0xbfa20002, 0x84788978, 4084 - 0xbfa00001, 0x84788a78, 4085 - 0xb8ee1e06, 0x846e8a6e, 4086 - 0x80786e78, 0x8078ff78, 4087 - 0x00000200, 0xbef600ff, 4088 - 0x01000000, 0xbeff0071, 4089 - 0xf4621bfa, 0xf0000000, 4090 - 0x80788478, 0xf4621b3a, 4091 - 0xf0000000, 0x80788478, 4092 - 0xf4621b7a, 0xf0000000, 4093 - 0x80788478, 0xf4621c3a, 4094 - 0xf0000000, 0x80788478, 4095 - 0xf4621c7a, 0xf0000000, 4096 - 0x80788478, 0xf4621eba, 4097 - 0xf0000000, 0x80788478, 4098 - 0xf4621efa, 0xf0000000, 4099 - 0x80788478, 0xf4621e7a, 4100 - 0xf0000000, 0x80788478, 4101 - 0xf4621cfa, 0xf0000000, 4102 - 0x80788478, 0xf4621bba, 4103 - 0xf0000000, 0x80788478, 4104 - 0xbf8a0000, 0xb96ef814, 4105 - 0xf4621bba, 0xf0000000, 4106 - 0x80788478, 0xbf8a0000, 4107 - 0xb96ef815, 0xf4621bba, 4108 - 0xf0000000, 0x80788478, 4109 - 0xbf8a0000, 0xb96ef812, 4110 - 0xf4621bba, 0xf0000000, 4111 - 0x80788478, 0xbf8a0000, 4112 - 0xb96ef813, 0x8b6eff7f, 4113 - 0x04000000, 0xbfa1000d, 4114 - 0x80788478, 0xf4621bba, 4115 - 0xf0000000, 0x80788478, 4116 - 0xbf8a0000, 0xbf0d806e, 4117 - 0xbfa10006, 0x856e906e, 4118 - 0x8b6e6e6e, 0xbfa10003, 4119 - 0xbe804ec1, 0x816ec16e, 4120 - 0xbfa0fffb, 0xbefd006f, 4121 - 0xbefe0070, 0xbeff0071, 4122 - 0xb97b2011, 0x857b867b, 4123 - 0xb97b0191, 0x857b827b, 4124 - 0xb97bba11, 0xb973f801, 4125 - 0xb8ee3b05, 0x806e816e, 4074 + 0xbefd00ff, 0x0000006c, 4075 + 0x80f89078, 0xf462403a, 4076 + 0xf0000000, 0xbf8a0000, 4077 + 0x80fd847d, 0xbf800000, 4078 + 0xbe804300, 0xbe824302, 4079 + 0x80f8a078, 0xf462603a, 4080 + 0xf0000000, 0xbf8a0000, 4081 + 0x80fd887d, 0xbf800000, 4082 + 0xbe804300, 0xbe824302, 4083 + 0xbe844304, 0xbe864306, 4084 + 0x80f8c078, 0xf462803a, 4085 + 0xf0000000, 0xbf8a0000, 4086 + 0x80fd907d, 0xbf800000, 4087 + 0xbe804300, 0xbe824302, 4088 + 0xbe844304, 0xbe864306, 4089 + 0xbe884308, 0xbe8a430a, 4090 + 0xbe8c430c, 0xbe8e430e, 4091 + 0xbf06807d, 0xbfa1fff0, 4092 + 0xb980f801, 0x00000000, 4093 + 0xb8f83b05, 0x80788178, 4126 4094 0xbf0d9972, 0xbfa20002, 4127 - 0x846e896e, 0xbfa00001, 4128 - 0x846e8a6e, 0xb8ef1e06, 4129 - 0x846f8a6f, 0x806e6f6e, 4130 - 0x806eff6e, 0x00000200, 4131 - 0x806e746e, 0x826f8075, 4132 - 0x8b6fff6f, 0x0000ffff, 4133 - 0xf4605c37, 0xf8000050, 4134 - 0xf4605d37, 0xf8000060, 4135 - 0xf4601e77, 0xf8000074, 4136 - 0xbf8a0000, 0x8b6dff6d, 4137 - 0x0000ffff, 0x8bfe7e7e, 4138 - 0x8bea6a6a, 0xb97af804, 4095 + 0x84788978, 0xbfa00001, 4096 + 0x84788a78, 0xb8ee1e06, 4097 + 0x846e8a6e, 0x80786e78, 4098 + 0x8078ff78, 0x00000200, 4099 + 0xbef600ff, 0x01000000, 4100 + 0xbeff0071, 0xf4621bfa, 4101 + 0xf0000000, 0x80788478, 4102 + 0xf4621b3a, 0xf0000000, 4103 + 0x80788478, 0xf4621b7a, 4104 + 0xf0000000, 0x80788478, 4105 + 0xf4621c3a, 0xf0000000, 4106 + 0x80788478, 0xf4621c7a, 4107 + 0xf0000000, 0x80788478, 4108 + 0xf4621eba, 0xf0000000, 4109 + 0x80788478, 0xf4621efa, 4110 + 0xf0000000, 0x80788478, 4111 + 0xf4621e7a, 0xf0000000, 4112 + 0x80788478, 0xf4621cfa, 4113 + 0xf0000000, 0x80788478, 4114 + 0xf4621bba, 0xf0000000, 4115 + 0x80788478, 0xbf8a0000, 4116 + 0xb96ef814, 0xf4621bba, 4117 + 0xf0000000, 0x80788478, 4118 + 0xbf8a0000, 0xb96ef815, 4119 + 0xf4621bba, 0xf0000000, 4120 + 0x80788478, 0xbf8a0000, 4121 + 0xb96ef812, 0xf4621bba, 4122 + 0xf0000000, 0x80788478, 4123 + 0xbf8a0000, 0xb96ef813, 4124 + 0x8b6eff7f, 0x04000000, 4125 + 0xbfa1000d, 0x80788478, 4126 + 0xf4621bba, 0xf0000000, 4127 + 0x80788478, 0xbf8a0000, 4128 + 0xbf0d806e, 0xbfa10006, 4129 + 0x856e906e, 0x8b6e6e6e, 4130 + 0xbfa10003, 0xbe804ec1, 4131 + 0x816ec16e, 0xbfa0fffb, 4132 + 0xbefd006f, 0xbefe0070, 4133 + 0xbeff0071, 0xb97b2011, 4134 + 0x857b867b, 0xb97b0191, 4135 + 0x857b827b, 0xb97bba11, 4136 + 0xb973f801, 0xb8ee3b05, 4137 + 0x806e816e, 0xbf0d9972, 4138 + 0xbfa20002, 0x846e896e, 4139 + 0xbfa00001, 0x846e8a6e, 4140 + 0xb8ef1e06, 0x846f8a6f, 4141 + 0x806e6f6e, 0x806eff6e, 4142 + 0x00000200, 0x806e746e, 4143 + 0x826f8075, 0x8b6fff6f, 4144 + 0x0000ffff, 0xf4605c37, 4145 + 0xf8000050, 0xf4605d37, 4146 + 0xf8000060, 0xf4601e77, 4147 + 0xf8000074, 0xbf8a0000, 4148 + 0x8b6dff6d, 0x0000ffff, 4149 + 0x8bfe7e7e, 0x8bea6a6a, 4150 + 0xb97af804, 0xbe804ec2, 4151 + 0xbf94fffe, 0xbe804a6c, 4139 4152 0xbe804ec2, 0xbf94fffe, 4140 - 0xbe804a6c, 0xbe804ec2, 4141 - 0xbf94fffe, 0xbfb10000, 4153 + 0xbfb10000, 0xbf9f0000, 4142 4154 0xbf9f0000, 0xbf9f0000, 4143 4155 0xbf9f0000, 0xbf9f0000, 4144 - 0xbf9f0000, 0x00000000, 4145 4156 }; 4146 4157 4147 4158 static const uint32_t cwsr_trap_gfx9_5_0_hex[] = {
+44 -38
drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx12.asm
··· 30 30 #define CHIP_GFX12 37 31 31 32 32 #define SINGLE_STEP_MISSED_WORKAROUND 1 //workaround for lost TRAP_AFTER_INST exception when SAVECTX raised 33 + #define HAVE_VALU_SGPR_HAZARD (ASIC_FAMILY == CHIP_GFX12) 33 34 34 35 var SQ_WAVE_STATE_PRIV_BARRIER_COMPLETE_MASK = 0x4 35 36 var SQ_WAVE_STATE_PRIV_SCC_SHIFT = 9 ··· 352 351 v_writelane_b32 v0, ttmp13, 0xD 353 352 v_writelane_b32 v0, exec_lo, 0xE 354 353 v_writelane_b32 v0, exec_hi, 0xF 354 + valu_sgpr_hazard() 355 355 356 356 s_mov_b32 exec_lo, 0x3FFF 357 357 s_mov_b32 exec_hi, 0x0 ··· 419 417 v_mov_b32 v0, 0x0 //Offset[31:0] from buffer resource 420 418 v_mov_b32 v1, 0x0 //Offset[63:32] from buffer resource 421 419 v_mov_b32 v2, 0x0 //Set of SGPRs for TCP store 422 - s_mov_b32 m0, 0x0 //Next lane of v2 to write to 423 420 424 421 // Ensure no further changes to barrier or LDS state. 425 422 // STATE_PRIV.BARRIER_COMPLETE may change up to this point. ··· 431 430 s_andn2_b32 s_save_state_priv, s_save_state_priv, SQ_WAVE_STATE_PRIV_BARRIER_COMPLETE_MASK 432 431 s_or_b32 s_save_state_priv, s_save_state_priv, s_save_tmp 433 432 434 - write_hwreg_to_v2(s_save_m0) 435 - write_hwreg_to_v2(s_save_pc_lo) 436 433 s_andn2_b32 s_save_tmp, s_save_pc_hi, S_SAVE_PC_HI_FIRST_WAVE_MASK 437 - write_hwreg_to_v2(s_save_tmp) 438 - write_hwreg_to_v2(s_save_exec_lo) 439 - write_hwreg_to_v2(s_save_exec_hi) 440 - write_hwreg_to_v2(s_save_state_priv) 434 + v_writelane_b32 v2, s_save_m0, 0x0 435 + v_writelane_b32 v2, s_save_pc_lo, 0x1 436 + v_writelane_b32 v2, s_save_tmp, 0x2 437 + v_writelane_b32 v2, s_save_exec_lo, 0x3 438 + v_writelane_b32 v2, s_save_exec_hi, 0x4 439 + v_writelane_b32 v2, s_save_state_priv, 0x5 440 + v_writelane_b32 v2, s_save_xnack_mask, 0x7 441 + valu_sgpr_hazard() 441 442 442 443 s_getreg_b32 s_save_tmp, hwreg(HW_REG_WAVE_EXCP_FLAG_PRIV) 443 - write_hwreg_to_v2(s_save_tmp) 444 + v_writelane_b32 v2, s_save_tmp, 0x6 444 445 445 - write_hwreg_to_v2(s_save_xnack_mask) 446 + s_getreg_b32 s_save_tmp, hwreg(HW_REG_WAVE_MODE) 447 + v_writelane_b32 v2, s_save_tmp, 0x8 446 448 447 - s_getreg_b32 s_save_m0, hwreg(HW_REG_WAVE_MODE) 448 - write_hwreg_to_v2(s_save_m0) 449 + s_getreg_b32 s_save_tmp, hwreg(HW_REG_WAVE_SCRATCH_BASE_LO) 450 + v_writelane_b32 v2, s_save_tmp, 0x9 449 451 450 - s_getreg_b32 s_save_m0, hwreg(HW_REG_WAVE_SCRATCH_BASE_LO) 451 - write_hwreg_to_v2(s_save_m0) 452 + s_getreg_b32 s_save_tmp, hwreg(HW_REG_WAVE_SCRATCH_BASE_HI) 453 + v_writelane_b32 v2, s_save_tmp, 0xA 452 454 453 - s_getreg_b32 s_save_m0, hwreg(HW_REG_WAVE_SCRATCH_BASE_HI) 454 - write_hwreg_to_v2(s_save_m0) 455 + s_getreg_b32 s_save_tmp, hwreg(HW_REG_WAVE_EXCP_FLAG_USER) 456 + v_writelane_b32 v2, s_save_tmp, 0xB 455 457 456 - s_getreg_b32 s_save_m0, hwreg(HW_REG_WAVE_EXCP_FLAG_USER) 457 - write_hwreg_to_v2(s_save_m0) 458 - 459 - s_getreg_b32 s_save_m0, hwreg(HW_REG_WAVE_TRAP_CTRL) 460 - write_hwreg_to_v2(s_save_m0) 458 + s_getreg_b32 s_save_tmp, hwreg(HW_REG_WAVE_TRAP_CTRL) 459 + v_writelane_b32 v2, s_save_tmp, 0xC 461 460 462 461 s_getreg_b32 s_save_tmp, hwreg(HW_REG_WAVE_STATUS) 463 - write_hwreg_to_v2(s_save_tmp) 462 + v_writelane_b32 v2, s_save_tmp, 0xD 464 463 465 464 s_get_barrier_state s_save_tmp, -1 466 465 s_wait_kmcnt (0) 467 - write_hwreg_to_v2(s_save_tmp) 466 + v_writelane_b32 v2, s_save_tmp, 0xE 467 + valu_sgpr_hazard() 468 468 469 469 // Write HWREGs with 16 VGPR lanes. TTMPs occupy space after this. 470 470 s_mov_b32 exec_lo, 0xFFFF ··· 499 497 s_movrels_b64 s12, s12 //s12 = s[12+m0], s13 = s[13+m0] 500 498 s_movrels_b64 s14, s14 //s14 = s[14+m0], s15 = s[15+m0] 501 499 502 - write_16sgpr_to_v2(s0) 503 - 504 - s_cmp_eq_u32 ttmp13, 0x20 //have 32 VGPR lanes filled? 505 - s_cbranch_scc0 L_SAVE_SGPR_SKIP_TCP_STORE 500 + s_cmp_eq_u32 ttmp13, 0x0 501 + s_cbranch_scc0 L_WRITE_V2_SECOND_HALF 502 + write_16sgpr_to_v2(s0, 0x0) 503 + s_branch L_SAVE_SGPR_SKIP_TCP_STORE 504 + L_WRITE_V2_SECOND_HALF: 505 + write_16sgpr_to_v2(s0, 0x10) 506 506 507 507 buffer_store_dword v2, v0, s_save_buf_rsrc0, s_save_mem_offset scope:SCOPE_SYS 508 508 s_add_u32 s_save_mem_offset, s_save_mem_offset, 0x80 ··· 1060 1056 s_endpgm_saved 1061 1057 end 1062 1058 1063 - function write_hwreg_to_v2(s) 1064 - // Copy into VGPR for later TCP store. 1065 - v_writelane_b32 v2, s, m0 1066 - s_add_u32 m0, m0, 0x1 1067 - end 1068 - 1069 - 1070 - function write_16sgpr_to_v2(s) 1059 + function write_16sgpr_to_v2(s, lane_offset) 1071 1060 // Copy into VGPR for later TCP store. 1072 1061 for var sgpr_idx = 0; sgpr_idx < 16; sgpr_idx ++ 1073 - v_writelane_b32 v2, s[sgpr_idx], ttmp13 1074 - s_add_u32 ttmp13, ttmp13, 0x1 1062 + v_writelane_b32 v2, s[sgpr_idx], sgpr_idx + lane_offset 1075 1063 end 1064 + valu_sgpr_hazard() 1065 + s_add_u32 ttmp13, ttmp13, 0x10 1076 1066 end 1077 1067 1078 1068 function write_12sgpr_to_v2(s) 1079 1069 // Copy into VGPR for later TCP store. 1080 1070 for var sgpr_idx = 0; sgpr_idx < 12; sgpr_idx ++ 1081 - v_writelane_b32 v2, s[sgpr_idx], ttmp13 1082 - s_add_u32 ttmp13, ttmp13, 0x1 1071 + v_writelane_b32 v2, s[sgpr_idx], sgpr_idx 1083 1072 end 1073 + valu_sgpr_hazard() 1084 1074 end 1085 1075 1086 1076 function read_hwreg_from_mem(s, s_rsrc, s_mem_offset) ··· 1125 1127 function get_wave_size2(s_reg) 1126 1128 s_getreg_b32 s_reg, hwreg(HW_REG_WAVE_STATUS,SQ_WAVE_STATUS_WAVE64_SHIFT,SQ_WAVE_STATUS_WAVE64_SIZE) 1127 1129 s_lshl_b32 s_reg, s_reg, S_WAVE_SIZE 1130 + end 1131 + 1132 + function valu_sgpr_hazard 1133 + #if HAVE_VALU_SGPR_HAZARD 1134 + for var rep = 0; rep < 8; rep ++ 1135 + ds_nop 1136 + end 1137 + #endif 1128 1138 end