/* * Copyright 2018 Advanced Micro Devices, Inc. * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), * to deal in the Software without restriction, including without limitation * the rights to use, copy, modify, merge, publish, distribute, sublicense, * and/or sell copies of the Software, and to permit persons to whom the * Software is furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included in * all copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR * OTHER DEALINGS IN THE SOFTWARE. */ shader main asic(DEFAULT) type(CS) wave_size(32) /*************************************************************************/ /* control on how to run the shader */ /*************************************************************************/ //any hack that needs to be made to run this code in EMU (either becasue various EMU code are not ready or no compute save & restore in EMU run) var EMU_RUN_HACK = 0 var EMU_RUN_HACK_RESTORE_NORMAL = 0 var EMU_RUN_HACK_SAVE_NORMAL_EXIT = 0 var EMU_RUN_HACK_SAVE_SINGLE_WAVE = 0 var EMU_RUN_HACK_SAVE_FIRST_TIME = 0 //for interrupted restore in which the first save is through EMU_RUN_HACK var SAVE_LDS = 0 var WG_BASE_ADDR_LO = 0x9000a000 var WG_BASE_ADDR_HI = 0x0 var WAVE_SPACE = 0x9000 //memory size that each wave occupies in workgroup state mem, increase from 5000 to 9000 for more SGPR need to be saved var CTX_SAVE_CONTROL = 0x0 var CTX_RESTORE_CONTROL = CTX_SAVE_CONTROL var SIM_RUN_HACK = 0 //any hack that needs to be made to run this code in SIM (either becasue various RTL code are not ready or no compute save & restore in RTL run) var SGPR_SAVE_USE_SQC = 0 //use SQC D$ to do the write var USE_MTBUF_INSTEAD_OF_MUBUF = 0 //need to change BUF_DATA_FORMAT in S_SAVE_BUF_RSRC_WORD3_MISC from 0 to BUF_DATA_FORMAT_32 if set to 1 (i.e. 0x00827FAC) var SWIZZLE_EN = 0 //whether we use swizzled buffer addressing var SAVE_RESTORE_HWID_DDID = 0 var RESTORE_DDID_IN_SGPR18 = 0 /**************************************************************************/ /* variables */ /**************************************************************************/ var SQ_WAVE_STATUS_INST_ATC_SHIFT = 23 var SQ_WAVE_STATUS_INST_ATC_MASK = 0x00800000 var SQ_WAVE_STATUS_SPI_PRIO_MASK = 0x00000006 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT = 12 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE = 9 var SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT = 8 var SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE = 6 var SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT = 24 var SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE = 4 //FIXME sq.blk still has 4 bits at this time while SQ programming guide has 3 bits var SQ_WAVE_LDS_ALLOC_VGPR_SHARED_SIZE_SHIFT = 24 var SQ_WAVE_LDS_ALLOC_VGPR_SHARED_SIZE_SIZE = 4 var SQ_WAVE_IB_STS2_WAVE64_SHIFT = 11 var SQ_WAVE_IB_STS2_WAVE64_SIZE = 1 var SQ_WAVE_TRAPSTS_SAVECTX_MASK = 0x400 var SQ_WAVE_TRAPSTS_EXCE_MASK = 0x1FF // Exception mask var SQ_WAVE_TRAPSTS_SAVECTX_SHIFT = 10 var SQ_WAVE_TRAPSTS_MEM_VIOL_MASK = 0x100 var SQ_WAVE_TRAPSTS_MEM_VIOL_SHIFT = 8 var SQ_WAVE_TRAPSTS_PRE_SAVECTX_MASK = 0x3FF var SQ_WAVE_TRAPSTS_PRE_SAVECTX_SHIFT = 0x0 var SQ_WAVE_TRAPSTS_PRE_SAVECTX_SIZE = 10 var SQ_WAVE_TRAPSTS_POST_SAVECTX_MASK = 0xFFFFF800 var SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT = 11 var SQ_WAVE_TRAPSTS_POST_SAVECTX_SIZE = 21 var SQ_WAVE_IB_STS_RCNT_SHIFT = 16 //FIXME var SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT = 15 //FIXME var SQ_WAVE_IB_STS_FIRST_REPLAY_SIZE = 1 //FIXME var SQ_WAVE_IB_STS_RCNT_SIZE = 6 //FIXME var SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK_NEG = 0x00007FFF //FIXME var SQ_BUF_RSRC_WORD1_ATC_SHIFT = 24 var SQ_BUF_RSRC_WORD3_MTYPE_SHIFT = 27 /* Save */ var S_SAVE_BUF_RSRC_WORD1_STRIDE = 0x00040000 //stride is 4 bytes var S_SAVE_BUF_RSRC_WORD3_MISC = 0x00807FAC //SQ_SEL_X/Y/Z/W, BUF_NUM_FORMAT_FLOAT, (0 for MUBUF stride[17:14] when ADD_TID_ENABLE and BUF_DATA_FORMAT_32 for MTBUF), ADD_TID_ENABLE var S_SAVE_SPI_INIT_ATC_MASK = 0x08000000 //bit[27]: ATC bit var S_SAVE_SPI_INIT_ATC_SHIFT = 27 var S_SAVE_SPI_INIT_MTYPE_MASK = 0x70000000 //bit[30:28]: Mtype var S_SAVE_SPI_INIT_MTYPE_SHIFT = 28 var S_SAVE_SPI_INIT_FIRST_WAVE_MASK = 0x04000000 //bit[26]: FirstWaveInTG var S_SAVE_SPI_INIT_FIRST_WAVE_SHIFT = 26 var S_SAVE_PC_HI_RCNT_SHIFT = 28 //FIXME check with Brian to ensure all fields other than PC[47:0] can be used var S_SAVE_PC_HI_RCNT_MASK = 0xF0000000 //FIXME var S_SAVE_PC_HI_FIRST_REPLAY_SHIFT = 27 //FIXME var S_SAVE_PC_HI_FIRST_REPLAY_MASK = 0x08000000 //FIXME var s_save_spi_init_lo = exec_lo var s_save_spi_init_hi = exec_hi var s_save_pc_lo = ttmp0 //{TTMP1, TTMP0} = {3¡¯h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]} var s_save_pc_hi = ttmp1 var s_save_exec_lo = ttmp2 var s_save_exec_hi = ttmp3 var s_save_status = ttmp4 var s_save_trapsts = ttmp5 //not really used until the end of the SAVE routine var s_wave_size = ttmp6 //ttmp6 is not needed now, since it's only 32bit xnack mask, now use it to determine wave32 or wave64 in EMU_HACK var s_save_xnack_mask = ttmp7 var s_save_buf_rsrc0 = ttmp8 var s_save_buf_rsrc1 = ttmp9 var s_save_buf_rsrc2 = ttmp10 var s_save_buf_rsrc3 = ttmp11 var s_save_mem_offset = ttmp14 var s_sgpr_save_num = 106 //in gfx10, all sgpr must be saved var s_save_alloc_size = s_save_trapsts //conflict var s_save_tmp = s_save_buf_rsrc2 //shared with s_save_buf_rsrc2 (conflict: should not use mem access with s_save_tmp at the same time) var s_save_m0 = ttmp15 /* Restore */ var S_RESTORE_BUF_RSRC_WORD1_STRIDE = S_SAVE_BUF_RSRC_WORD1_STRIDE var S_RESTORE_BUF_RSRC_WORD3_MISC = S_SAVE_BUF_RSRC_WORD3_MISC var S_RESTORE_SPI_INIT_ATC_MASK = 0x08000000 //bit[27]: ATC bit var S_RESTORE_SPI_INIT_ATC_SHIFT = 27 var S_RESTORE_SPI_INIT_MTYPE_MASK = 0x70000000 //bit[30:28]: Mtype var S_RESTORE_SPI_INIT_MTYPE_SHIFT = 28 var S_RESTORE_SPI_INIT_FIRST_WAVE_MASK = 0x04000000 //bit[26]: FirstWaveInTG var S_RESTORE_SPI_INIT_FIRST_WAVE_SHIFT = 26 var S_RESTORE_PC_HI_RCNT_SHIFT = S_SAVE_PC_HI_RCNT_SHIFT var S_RESTORE_PC_HI_RCNT_MASK = S_SAVE_PC_HI_RCNT_MASK var S_RESTORE_PC_HI_FIRST_REPLAY_SHIFT = S_SAVE_PC_HI_FIRST_REPLAY_SHIFT var S_RESTORE_PC_HI_FIRST_REPLAY_MASK = S_SAVE_PC_HI_FIRST_REPLAY_MASK var s_restore_spi_init_lo = exec_lo var s_restore_spi_init_hi = exec_hi var s_restore_mem_offset = ttmp12 var s_restore_alloc_size = ttmp3 var s_restore_tmp = ttmp6 var s_restore_mem_offset_save = s_restore_tmp //no conflict var s_restore_m0 = s_restore_alloc_size //no conflict var s_restore_mode = ttmp13 var s_restore_hwid1 = ttmp2 var s_restore_ddid = s_restore_hwid1 var s_restore_pc_lo = ttmp0 var s_restore_pc_hi = ttmp1 var s_restore_exec_lo = ttmp14 var s_restore_exec_hi = ttmp15 var s_restore_status = ttmp4 var s_restore_trapsts = ttmp5 //var s_restore_xnack_mask_lo = xnack_mask_lo //var s_restore_xnack_mask_hi = xnack_mask_hi var s_restore_xnack_mask = ttmp7 var s_restore_buf_rsrc0 = ttmp8 var s_restore_buf_rsrc1 = ttmp9 var s_restore_buf_rsrc2 = ttmp10 var s_restore_buf_rsrc3 = ttmp11 var s_restore_size = ttmp13 //ttmp13 has no conflict /**************************************************************************/ /* trap handler entry points */ /**************************************************************************/ if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL)) //hack to use trap_id for determining save/restore //FIXME VCCZ un-init assertion s_getreg_b32 s_save_status, hwreg(HW_REG_STATUS) //save STATUS since we will change SCC s_and_b32 s_save_tmp, s_save_pc_hi, 0xffff0000 //change SCC s_cmp_eq_u32 s_save_tmp, 0x007e0000 //Save: trap_id = 0x7e. Restore: trap_id = 0x7f. s_cbranch_scc0 L_JUMP_TO_RESTORE //do not need to recover STATUS here since we are going to RESTORE //FIXME s_setreg_b32 hwreg(HW_REG_STATUS), s_save_status //need to recover STATUS since we are going to SAVE s_branch L_SKIP_RESTORE //NOT restore, SAVE actually else s_branch L_SKIP_RESTORE //NOT restore. might be a regular trap or save end L_JUMP_TO_RESTORE: s_branch L_RESTORE //restore L_SKIP_RESTORE: s_getreg_b32 s_save_status, hwreg(HW_REG_STATUS) //save STATUS since we will change SCC s_andn2_b32 s_save_status, s_save_status, SQ_WAVE_STATUS_SPI_PRIO_MASK //check whether this is for save s_getreg_b32 s_save_trapsts, hwreg(HW_REG_TRAPSTS) s_and_b32 s_save_trapsts, s_save_trapsts, SQ_WAVE_TRAPSTS_SAVECTX_MASK //check whether this is for save s_cbranch_scc1 L_SAVE //this is the operation for save // ********* Handle non-CWSR traps ******************* if (!EMU_RUN_HACK) s_getreg_b32 s_save_trapsts, hwreg(HW_REG_TRAPSTS) s_and_b32 s_save_trapsts, s_save_trapsts, SQ_WAVE_TRAPSTS_EXCE_MASK // Check whether it is an exception s_cbranch_scc1 L_EXCP_CASE // Exception, jump back to the shader program directly. s_add_u32 ttmp0, ttmp0, 4 // S_TRAP case, add 4 to ttmp0 L_EXCP_CASE: s_and_b32 ttmp1, ttmp1, 0xFFFF s_rfe_b64 [ttmp0, ttmp1] end // ********* End handling of non-CWSR traps ******************* /**************************************************************************/ /* save routine */ /**************************************************************************/ L_SAVE: //check whether there is mem_viol s_getreg_b32 s_save_trapsts, hwreg(HW_REG_TRAPSTS) s_and_b32 s_save_trapsts, s_save_trapsts, SQ_WAVE_TRAPSTS_MEM_VIOL_MASK s_cbranch_scc0 L_NO_PC_REWIND //if so, need rewind PC assuming GDS operation gets NACKed s_mov_b32 s_save_tmp, 0 //clear mem_viol bit s_setreg_b32 hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_MEM_VIOL_SHIFT, 1), s_save_tmp //clear mem_viol bit s_and_b32 s_save_pc_hi, s_save_pc_hi, 0x0000ffff //pc[47:32] s_sub_u32 s_save_pc_lo, s_save_pc_lo, 8 //pc[31:0]-8 s_subb_u32 s_save_pc_hi, s_save_pc_hi, 0x0 // -scc L_NO_PC_REWIND: s_mov_b32 s_save_tmp, 0 //clear saveCtx bit s_setreg_b32 hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_SAVECTX_SHIFT, 1), s_save_tmp //clear saveCtx bit //s_mov_b32 s_save_xnack_mask_lo, xnack_mask_lo //save XNACK_MASK //s_mov_b32 s_save_xnack_mask_hi, xnack_mask_hi s_getreg_b32 s_save_xnack_mask, hwreg(HW_REG_SHADER_XNACK_MASK) s_getreg_b32 s_save_tmp, hwreg(HW_REG_IB_STS, SQ_WAVE_IB_STS_RCNT_SHIFT, SQ_WAVE_IB_STS_RCNT_SIZE) //save RCNT s_lshl_b32 s_save_tmp, s_save_tmp, S_SAVE_PC_HI_RCNT_SHIFT s_or_b32 s_save_pc_hi, s_save_pc_hi, s_save_tmp s_getreg_b32 s_save_tmp, hwreg(HW_REG_IB_STS, SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT, SQ_WAVE_IB_STS_FIRST_REPLAY_SIZE) //save FIRST_REPLAY s_lshl_b32 s_save_tmp, s_save_tmp, S_SAVE_PC_HI_FIRST_REPLAY_SHIFT s_or_b32 s_save_pc_hi, s_save_pc_hi, s_save_tmp s_getreg_b32 s_save_tmp, hwreg(HW_REG_IB_STS) //clear RCNT and FIRST_REPLAY in IB_STS s_and_b32 s_save_tmp, s_save_tmp, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK_NEG s_setreg_b32 hwreg(HW_REG_IB_STS), s_save_tmp /* inform SPI the readiness and wait for SPI's go signal */ s_mov_b32 s_save_exec_lo, exec_lo //save EXEC and use EXEC for the go signal from SPI s_mov_b32 s_save_exec_hi, exec_hi s_mov_b64 exec, 0x0 //clear EXEC to get ready to receive if (EMU_RUN_HACK) else s_sendmsg sendmsg(MSG_SAVEWAVE) //send SPI a message and wait for SPI's write to EXEC end L_SLEEP: s_sleep 0x2 if (EMU_RUN_HACK) else s_cbranch_execz L_SLEEP end /* setup Resource Contants */ if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_SAVE_SINGLE_WAVE)) //calculate wd_addr using absolute thread id v_readlane_b32 s_save_tmp, v9, 0 //determine it is wave32 or wave64 s_getreg_b32 s_wave_size, hwreg(HW_REG_IB_STS2,SQ_WAVE_IB_STS2_WAVE64_SHIFT,SQ_WAVE_IB_STS2_WAVE64_SIZE) s_cmp_eq_u32 s_wave_size, 0 s_cbranch_scc1 L_SAVE_WAVE32 s_lshr_b32 s_save_tmp, s_save_tmp, 6 //SAVE WAVE64 s_branch L_SAVE_CON L_SAVE_WAVE32: s_lshr_b32 s_save_tmp, s_save_tmp, 5 //SAVE WAVE32 L_SAVE_CON: s_mul_i32 s_save_tmp, s_save_tmp, WAVE_SPACE s_add_i32 s_save_spi_init_lo, s_save_tmp, WG_BASE_ADDR_LO s_mov_b32 s_save_spi_init_hi, WG_BASE_ADDR_HI s_and_b32 s_save_spi_init_hi, s_save_spi_init_hi, CTX_SAVE_CONTROL else end if ((EMU_RUN_HACK) && (EMU_RUN_HACK_SAVE_SINGLE_WAVE)) s_add_i32 s_save_spi_init_lo, s_save_tmp, WG_BASE_ADDR_LO s_mov_b32 s_save_spi_init_hi, WG_BASE_ADDR_HI s_and_b32 s_save_spi_init_hi, s_save_spi_init_hi, CTX_SAVE_CONTROL else end s_mov_b32 s_save_buf_rsrc0, s_save_spi_init_lo //base_addr_lo s_and_b32 s_save_buf_rsrc1, s_save_spi_init_hi, 0x0000FFFF //base_addr_hi s_or_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, S_SAVE_BUF_RSRC_WORD1_STRIDE s_mov_b32 s_save_buf_rsrc2, 0 //NUM_RECORDS initial value = 0 (in bytes) although not neccessarily inited s_mov_b32 s_save_buf_rsrc3, S_SAVE_BUF_RSRC_WORD3_MISC s_and_b32 s_save_tmp, s_save_spi_init_hi, S_SAVE_SPI_INIT_ATC_MASK s_lshr_b32 s_save_tmp, s_save_tmp, (S_SAVE_SPI_INIT_ATC_SHIFT-SQ_BUF_RSRC_WORD1_ATC_SHIFT) //get ATC bit into position s_or_b32 s_save_buf_rsrc3, s_save_buf_rsrc3, s_save_tmp //or ATC s_and_b32 s_save_tmp, s_save_spi_init_hi, S_SAVE_SPI_INIT_MTYPE_MASK s_lshr_b32 s_save_tmp, s_save_tmp, (S_SAVE_SPI_INIT_MTYPE_SHIFT-SQ_BUF_RSRC_WORD3_MTYPE_SHIFT) //get MTYPE bits into position s_or_b32 s_save_buf_rsrc3, s_save_buf_rsrc3, s_save_tmp //or MTYPE s_mov_b32 s_save_m0, m0 //save M0 /* global mem offset */ s_mov_b32 s_save_mem_offset, 0x0 //mem offset initial value = 0 s_getreg_b32 s_wave_size, hwreg(HW_REG_IB_STS2,SQ_WAVE_IB_STS2_WAVE64_SHIFT,SQ_WAVE_IB_STS2_WAVE64_SIZE) //get wave_save_size s_or_b32 s_wave_size, s_save_spi_init_hi, s_wave_size //share s_wave_size with exec_hi /* save VGPRs */ ////////////////////////////// L_SAVE_VGPR: s_mov_b32 exec_lo, 0xFFFFFFFF //need every thread from now on s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_ENABLE_SAVE_VGPR_EXEC_HI s_mov_b32 exec_hi, 0x00000000 s_branch L_SAVE_VGPR_NORMAL L_ENABLE_SAVE_VGPR_EXEC_HI: s_mov_b32 exec_hi, 0xFFFFFFFF L_SAVE_VGPR_NORMAL: s_getreg_b32 s_save_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE) //vpgr_size //for wave32 and wave64, the num of vgpr function is the same? s_add_u32 s_save_alloc_size, s_save_alloc_size, 1 s_lshl_b32 s_save_alloc_size, s_save_alloc_size, 2 //Number of VGPRs = (vgpr_size + 1) * 4 (non-zero value) //FIXME for GFX, zero is possible //determine it is wave32 or wave64 s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_SAVE_VGPR_WAVE64 //zhenxu added it for save vgpr for wave32 s_lshl_b32 s_save_buf_rsrc2, s_save_alloc_size, 7 //NUM_RECORDS in bytes (32 threads*4) if (SWIZZLE_EN) s_add_u32 s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_save_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_mov_b32 m0, 0x0 //VGPR initial index value =0 //s_set_gpr_idx_on m0, 0x1 //M0[7:0] = M0[7:0] and M0[15:12] = 0x1 //s_add_u32 s_save_alloc_size, s_save_alloc_size, 0x1000 //add 0x1000 since we compare m0 against it later, doesn't need this in gfx10 L_SAVE_VGPR_WAVE32_LOOP: v_movrels_b32 v0, v0 //v0 = v[0+m0] if(USE_MTBUF_INSTEAD_OF_MUBUF) tbuffer_store_format_x v0, v0, s_save_buf_rsrc0, s_save_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 else buffer_store_dword v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1 end s_add_u32 m0, m0, 1 //next vgpr index s_add_u32 s_save_mem_offset, s_save_mem_offset, 128 //every buffer_store_dword does 128 bytes s_cmp_lt_u32 m0, s_save_alloc_size //scc = (m0 < s_save_alloc_size) ? 1 : 0 s_cbranch_scc1 L_SAVE_VGPR_WAVE32_LOOP //VGPR save is complete? s_branch L_SAVE_LDS //save vgpr for wave32 ends L_SAVE_VGPR_WAVE64: s_lshl_b32 s_save_buf_rsrc2, s_save_alloc_size, 8 //NUM_RECORDS in bytes (64 threads*4) if (SWIZZLE_EN) s_add_u32 s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_save_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_mov_b32 m0, 0x0 //VGPR initial index value =0 //s_set_gpr_idx_on m0, 0x1 //M0[7:0] = M0[7:0] and M0[15:12] = 0x1 //s_add_u32 s_save_alloc_size, s_save_alloc_size, 0x1000 //add 0x1000 since we compare m0 against it later, doesn't need this in gfx10 L_SAVE_VGPR_WAVE64_LOOP: v_movrels_b32 v0, v0 //v0 = v[0+m0] if(USE_MTBUF_INSTEAD_OF_MUBUF) tbuffer_store_format_x v0, v0, s_save_buf_rsrc0, s_save_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 else buffer_store_dword v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1 end s_add_u32 m0, m0, 1 //next vgpr index s_add_u32 s_save_mem_offset, s_save_mem_offset, 256 //every buffer_store_dword does 256 bytes s_cmp_lt_u32 m0, s_save_alloc_size //scc = (m0 < s_save_alloc_size) ? 1 : 0 s_cbranch_scc1 L_SAVE_VGPR_WAVE64_LOOP //VGPR save is complete? //s_set_gpr_idx_off // //Below part will be the save shared vgpr part (new for gfx10) s_getreg_b32 s_save_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_VGPR_SHARED_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_VGPR_SHARED_SIZE_SIZE) //shared_vgpr_size s_and_b32 s_save_alloc_size, s_save_alloc_size, 0xFFFFFFFF //shared_vgpr_size is zero? s_cbranch_scc0 L_SAVE_LDS //no shared_vgpr used? jump to L_SAVE_LDS s_lshl_b32 s_save_alloc_size, s_save_alloc_size, 3 //Number of SHARED_VGPRs = shared_vgpr_size * 8 (non-zero value) //m0 now has the value of normal vgpr count, just add the m0 with shared_vgpr count to get the total count. //save shared_vgpr will start from the index of m0 s_add_u32 s_save_alloc_size, s_save_alloc_size, m0 s_mov_b32 exec_lo, 0xFFFFFFFF s_mov_b32 exec_hi, 0x00000000 L_SAVE_SHARED_VGPR_WAVE64_LOOP: v_movrels_b32 v0, v0 //v0 = v[0+m0] buffer_store_dword v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1 s_add_u32 m0, m0, 1 //next vgpr index s_add_u32 s_save_mem_offset, s_save_mem_offset, 128 //every buffer_store_dword does 256 bytes s_cmp_lt_u32 m0, s_save_alloc_size //scc = (m0 < s_save_alloc_size) ? 1 : 0 s_cbranch_scc1 L_SAVE_SHARED_VGPR_WAVE64_LOOP //SHARED_VGPR save is complete? /* save LDS */ ////////////////////////////// L_SAVE_LDS: //Only check the first wave need LDS /* the first wave in the threadgroup */ s_barrier //FIXME not performance-optimal "LDS is used? wait for other waves in the same TG" s_and_b32 s_save_tmp, s_wave_size, S_SAVE_SPI_INIT_FIRST_WAVE_MASK //exec is still used here s_cbranch_scc0 L_SAVE_SGPR s_mov_b32 exec_lo, 0xFFFFFFFF //need every thread from now on s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_ENABLE_SAVE_LDS_EXEC_HI s_mov_b32 exec_hi, 0x00000000 s_branch L_SAVE_LDS_NORMAL L_ENABLE_SAVE_LDS_EXEC_HI: s_mov_b32 exec_hi, 0xFFFFFFFF L_SAVE_LDS_NORMAL: s_getreg_b32 s_save_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE) //lds_size s_and_b32 s_save_alloc_size, s_save_alloc_size, 0xFFFFFFFF //lds_size is zero? s_cbranch_scc0 L_SAVE_SGPR //no lds used? jump to L_SAVE_VGPR s_lshl_b32 s_save_alloc_size, s_save_alloc_size, 6 //LDS size in dwords = lds_size * 64dw s_lshl_b32 s_save_alloc_size, s_save_alloc_size, 2 //LDS size in bytes s_mov_b32 s_save_buf_rsrc2, s_save_alloc_size //NUM_RECORDS in bytes if (SWIZZLE_EN) s_add_u32 s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_save_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end //load 0~63*4(byte address) to vgpr v15 v_mbcnt_lo_u32_b32 v0, -1, 0 v_mbcnt_hi_u32_b32 v0, -1, v0 v_mul_u32_u24 v0, 4, v0 s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_mov_b32 m0, 0x0 s_cbranch_scc1 L_SAVE_LDS_LOOP_W64 L_SAVE_LDS_LOOP_W32: if (SAVE_LDS) ds_read_b32 v1, v0 s_waitcnt 0 //ensure data ready buffer_store_dword v1, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1 //buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 //save lds to memory doesn't exist in 10 end s_add_u32 m0, m0, 128 //every buffer_store_lds does 128 bytes s_add_u32 s_save_mem_offset, s_save_mem_offset, 128 //mem offset increased by 128 bytes v_add_nc_u32 v0, v0, 128 s_cmp_lt_u32 m0, s_save_alloc_size //scc=(m0 < s_save_alloc_size) ? 1 : 0 s_cbranch_scc1 L_SAVE_LDS_LOOP_W32 //LDS save is complete? s_branch L_SAVE_SGPR L_SAVE_LDS_LOOP_W64: if (SAVE_LDS) ds_read_b32 v1, v0 s_waitcnt 0 //ensure data ready buffer_store_dword v1, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1 //buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 //save lds to memory doesn't exist in 10 end s_add_u32 m0, m0, 256 //every buffer_store_lds does 256 bytes s_add_u32 s_save_mem_offset, s_save_mem_offset, 256 //mem offset increased by 256 bytes v_add_nc_u32 v0, v0, 256 s_cmp_lt_u32 m0, s_save_alloc_size //scc=(m0 < s_save_alloc_size) ? 1 : 0 s_cbranch_scc1 L_SAVE_LDS_LOOP_W64 //LDS save is complete? /* save SGPRs */ ////////////////////////////// //s_getreg_b32 s_save_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE) //spgr_size //s_add_u32 s_save_alloc_size, s_save_alloc_size, 1 //s_lshl_b32 s_save_alloc_size, s_save_alloc_size, 4 //Number of SGPRs = (sgpr_size + 1) * 16 (non-zero value) //s_lshl_b32 s_save_alloc_size, s_save_alloc_size, 3 //In gfx10, Number of SGPRs = (sgpr_size + 1) * 8 (non-zero value) L_SAVE_SGPR: //need to look at it is wave32 or wave64 s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_SAVE_SGPR_VMEM_WAVE64 if (SGPR_SAVE_USE_SQC) s_lshl_b32 s_save_buf_rsrc2, s_sgpr_save_num, 2 //NUM_RECORDS in bytes else s_lshl_b32 s_save_buf_rsrc2, s_sgpr_save_num, 7 //NUM_RECORDS in bytes (32 threads) end s_branch L_SAVE_SGPR_CONT L_SAVE_SGPR_VMEM_WAVE64: if (SGPR_SAVE_USE_SQC) s_lshl_b32 s_save_buf_rsrc2, s_sgpr_save_num, 2 //NUM_RECORDS in bytes else s_lshl_b32 s_save_buf_rsrc2, s_sgpr_save_num, 8 //NUM_RECORDS in bytes (64 threads) end L_SAVE_SGPR_CONT: if (SWIZZLE_EN) s_add_u32 s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_save_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end //s_mov_b32 m0, 0x0 //SGPR initial index value =0 //s_nop 0x0 //Manually inserted wait states s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_mov_b32 m0, 0x0 //SGPR initial index value =0 s_nop 0x0 //Manually inserted wait states s_cbranch_scc1 L_SAVE_SGPR_LOOP_WAVE64 L_SAVE_SGPR_LOOP_WAVE32: s_movrels_b32 s0, s0 //s0 = s[0+m0] //zhenxu, adding one more argument to save sgpr function, this is only for vmem, using sqc is not change write_sgpr_to_mem_wave32(s0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //PV: the best performance should be using s_buffer_store_dwordx4 s_add_u32 m0, m0, 1 //next sgpr index s_cmp_lt_u32 m0, s_sgpr_save_num //scc = (m0 < s_sgpr_save_num) ? 1 : 0 s_cbranch_scc1 L_SAVE_SGPR_LOOP_WAVE32 //SGPR save is complete? s_branch L_SAVE_HWREG L_SAVE_SGPR_LOOP_WAVE64: s_movrels_b32 s0, s0 //s0 = s[0+m0] //zhenxu, adding one more argument to save sgpr function, this is only for vmem, using sqc is not change write_sgpr_to_mem_wave64(s0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //PV: the best performance should be using s_buffer_store_dwordx4 s_add_u32 m0, m0, 1 //next sgpr index s_cmp_lt_u32 m0, s_sgpr_save_num //scc = (m0 < s_sgpr_save_num) ? 1 : 0 s_cbranch_scc1 L_SAVE_SGPR_LOOP_WAVE64 //SGPR save is complete? /* save HW registers */ ////////////////////////////// L_SAVE_HWREG: s_mov_b32 s_save_buf_rsrc2, 0x4 //NUM_RECORDS in bytes if (SWIZZLE_EN) s_add_u32 s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_save_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_SAVE_HWREG_WAVE64 write_sgpr_to_mem_wave32(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //M0 if ((EMU_RUN_HACK) && (EMU_RUN_HACK_SAVE_FIRST_TIME)) s_add_u32 s_save_pc_lo, s_save_pc_lo, 4 //pc[31:0]+4 s_addc_u32 s_save_pc_hi, s_save_pc_hi, 0x0 //carry bit over end write_sgpr_to_mem_wave32(s_save_pc_lo, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //PC write_sgpr_to_mem_wave32(s_save_pc_hi, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) write_sgpr_to_mem_wave32(s_save_exec_lo, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //EXEC write_sgpr_to_mem_wave32(s_save_exec_hi, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) write_sgpr_to_mem_wave32(s_save_status, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //STATUS //s_save_trapsts conflicts with s_save_alloc_size s_getreg_b32 s_save_trapsts, hwreg(HW_REG_TRAPSTS) write_sgpr_to_mem_wave32(s_save_trapsts, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //TRAPSTS //write_sgpr_to_mem_wave32(s_save_xnack_mask_lo, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //XNACK_MASK_LO write_sgpr_to_mem_wave32(s_save_xnack_mask, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //XNACK_MASK_HI //use s_save_tmp would introduce conflict here between s_save_tmp and s_save_buf_rsrc2 s_getreg_b32 s_save_m0, hwreg(HW_REG_MODE) //MODE write_sgpr_to_mem_wave32(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) if(SAVE_RESTORE_HWID_DDID) s_getreg_b32 s_save_m0, hwreg(HW_REG_HW_ID1) //HW_ID1, handler records the SE/SA/WGP/SIMD/wave of the original wave write_sgpr_to_mem_wave32(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) end s_branch L_S_PGM_END_SAVED L_SAVE_HWREG_WAVE64: write_sgpr_to_mem_wave64(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //M0 if ((EMU_RUN_HACK) && (EMU_RUN_HACK_SAVE_FIRST_TIME)) s_add_u32 s_save_pc_lo, s_save_pc_lo, 4 //pc[31:0]+4 s_addc_u32 s_save_pc_hi, s_save_pc_hi, 0x0 //carry bit over end write_sgpr_to_mem_wave64(s_save_pc_lo, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //PC write_sgpr_to_mem_wave64(s_save_pc_hi, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) write_sgpr_to_mem_wave64(s_save_exec_lo, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //EXEC write_sgpr_to_mem_wave64(s_save_exec_hi, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) write_sgpr_to_mem_wave64(s_save_status, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //STATUS //s_save_trapsts conflicts with s_save_alloc_size s_getreg_b32 s_save_trapsts, hwreg(HW_REG_TRAPSTS) write_sgpr_to_mem_wave64(s_save_trapsts, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //TRAPSTS //write_sgpr_to_mem_wave64(s_save_xnack_mask_lo, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //XNACK_MASK_LO write_sgpr_to_mem_wave64(s_save_xnack_mask, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) //XNACK_MASK_HI //use s_save_tmp would introduce conflict here between s_save_tmp and s_save_buf_rsrc2 s_getreg_b32 s_save_m0, hwreg(HW_REG_MODE) //MODE write_sgpr_to_mem_wave64(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) if(SAVE_RESTORE_HWID_DDID) s_getreg_b32 s_save_m0, hwreg(HW_REG_HW_ID1) //HW_ID1, handler records the SE/SA/WGP/SIMD/wave of the original wave write_sgpr_to_mem_wave64(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) /* save DDID */ ////////////////////////////// L_SAVE_DDID: //EXEC has been saved, no vector inst following s_mov_b32 exec_lo, 0x80000000 //Set MSB to 1. Cleared when draw index is returned s_sendmsg sendmsg(MSG_GET_DDID) L_WAIT_DDID_LOOP: s_nop 7 // sleep a bit s_bitcmp0_b32 exec_lo, 31 // test to see if MSB is cleared, meaning done s_cbranch_scc0 L_WAIT_DDID_LOOP s_mov_b32 s_save_m0, exec_lo s_mov_b32 s_save_buf_rsrc2, 0x4 //NUM_RECORDS in bytes if (SWIZZLE_EN) s_add_u32 s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_save_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_SAVE_DDID_WAVE64 write_sgpr_to_mem_wave32(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) L_SAVE_DDID_WAVE64: write_sgpr_to_mem_wave64(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset, SGPR_SAVE_USE_SQC, USE_MTBUF_INSTEAD_OF_MUBUF) end L_S_PGM_END_SAVED: /* S_PGM_END_SAVED */ //FIXME graphics ONLY if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_SAVE_NORMAL_EXIT)) s_and_b32 s_save_pc_hi, s_save_pc_hi, 0x0000ffff //pc[47:32] s_add_u32 s_save_pc_lo, s_save_pc_lo, 4 //pc[31:0]+4 s_addc_u32 s_save_pc_hi, s_save_pc_hi, 0x0 //carry bit over s_rfe_b64 s_save_pc_lo //Return to the main shader program else end s_branch L_END_PGM /**************************************************************************/ /* restore routine */ /**************************************************************************/ L_RESTORE: /* Setup Resource Contants */ if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL)) //calculate wd_addr using absolute thread id v_readlane_b32 s_restore_tmp, v9, 0 //determine it is wave32 or wave64 s_getreg_b32 s_restore_size, hwreg(HW_REG_IB_STS2,SQ_WAVE_IB_STS2_WAVE64_SHIFT,SQ_WAVE_IB_STS2_WAVE64_SIZE) //change to ttmp13 s_cmp_eq_u32 s_restore_size, 0 s_cbranch_scc1 L_RESTORE_WAVE32 s_lshr_b32 s_restore_tmp, s_restore_tmp, 6 //SAVE WAVE64 s_branch L_RESTORE_CON L_RESTORE_WAVE32: s_lshr_b32 s_restore_tmp, s_restore_tmp, 5 //SAVE WAVE32 L_RESTORE_CON: s_mul_i32 s_restore_tmp, s_restore_tmp, WAVE_SPACE s_add_i32 s_restore_spi_init_lo, s_restore_tmp, WG_BASE_ADDR_LO s_mov_b32 s_restore_spi_init_hi, WG_BASE_ADDR_HI s_and_b32 s_restore_spi_init_hi, s_restore_spi_init_hi, CTX_RESTORE_CONTROL else end s_mov_b32 s_restore_buf_rsrc0, s_restore_spi_init_lo //base_addr_lo s_and_b32 s_restore_buf_rsrc1, s_restore_spi_init_hi, 0x0000FFFF //base_addr_hi s_or_b32 s_restore_buf_rsrc1, s_restore_buf_rsrc1, S_RESTORE_BUF_RSRC_WORD1_STRIDE s_mov_b32 s_restore_buf_rsrc2, 0 //NUM_RECORDS initial value = 0 (in bytes) s_mov_b32 s_restore_buf_rsrc3, S_RESTORE_BUF_RSRC_WORD3_MISC s_and_b32 s_restore_tmp, s_restore_spi_init_hi, S_RESTORE_SPI_INIT_ATC_MASK s_lshr_b32 s_restore_tmp, s_restore_tmp, (S_RESTORE_SPI_INIT_ATC_SHIFT-SQ_BUF_RSRC_WORD1_ATC_SHIFT) //get ATC bit into position s_or_b32 s_restore_buf_rsrc3, s_restore_buf_rsrc3, s_restore_tmp //or ATC s_and_b32 s_restore_tmp, s_restore_spi_init_hi, S_RESTORE_SPI_INIT_MTYPE_MASK s_lshr_b32 s_restore_tmp, s_restore_tmp, (S_RESTORE_SPI_INIT_MTYPE_SHIFT-SQ_BUF_RSRC_WORD3_MTYPE_SHIFT) //get MTYPE bits into position s_or_b32 s_restore_buf_rsrc3, s_restore_buf_rsrc3, s_restore_tmp //or MTYPE //determine it is wave32 or wave64 s_getreg_b32 s_restore_size, hwreg(HW_REG_IB_STS2,SQ_WAVE_IB_STS2_WAVE64_SHIFT,SQ_WAVE_IB_STS2_WAVE64_SIZE) s_or_b32 s_restore_size, s_restore_spi_init_hi, s_restore_size //share s_wave_size with exec_hi /* global mem offset */ s_mov_b32 s_restore_mem_offset, 0x0 //mem offset initial value = 0 /* restore VGPRs */ ////////////////////////////// L_RESTORE_VGPR: s_mov_b32 exec_lo, 0xFFFFFFFF //need every thread from now on //be consistent with SAVE although can be moved ahead s_and_b32 m0, s_restore_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_ENABLE_RESTORE_VGPR_EXEC_HI s_mov_b32 exec_hi, 0x00000000 s_branch L_RESTORE_VGPR_NORMAL L_ENABLE_RESTORE_VGPR_EXEC_HI: s_mov_b32 exec_hi, 0xFFFFFFFF L_RESTORE_VGPR_NORMAL: s_getreg_b32 s_restore_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE) //vpgr_size s_add_u32 s_restore_alloc_size, s_restore_alloc_size, 1 s_lshl_b32 s_restore_alloc_size, s_restore_alloc_size, 2 //Number of VGPRs = (vgpr_size + 1) * 4 (non-zero value) //determine it is wave32 or wave64 s_and_b32 m0, s_restore_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_RESTORE_VGPR_WAVE64 s_lshl_b32 s_restore_buf_rsrc2, s_restore_alloc_size, 7 //NUM_RECORDS in bytes (32 threads*4) if (SWIZZLE_EN) s_add_u32 s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_restore_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_mov_b32 s_restore_mem_offset_save, s_restore_mem_offset // restore start with v1, v0 will be the last s_add_u32 s_restore_mem_offset, s_restore_mem_offset, 128 s_mov_b32 m0, 1 //VGPR initial index value = 1 //s_set_gpr_idx_on m0, 0x8 //M0[7:0] = M0[7:0] and M0[15:12] = 0x8 //s_add_u32 s_restore_alloc_size, s_restore_alloc_size, 0x8000 //add 0x8000 since we compare m0 against it later, might not need this in gfx10 L_RESTORE_VGPR_WAVE32_LOOP: if(USE_MTBUF_INSTEAD_OF_MUBUF) tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 else buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 end s_waitcnt vmcnt(0) //ensure data ready v_movreld_b32 v0, v0 //v[0+m0] = v0 s_add_u32 m0, m0, 1 //next vgpr index s_add_u32 s_restore_mem_offset, s_restore_mem_offset, 128 //every buffer_load_dword does 128 bytes s_cmp_lt_u32 m0, s_restore_alloc_size //scc = (m0 < s_restore_alloc_size) ? 1 : 0 s_cbranch_scc1 L_RESTORE_VGPR_WAVE32_LOOP //VGPR restore (except v0) is complete? //s_set_gpr_idx_off /* VGPR restore on v0 */ if(USE_MTBUF_INSTEAD_OF_MUBUF) tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 else buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save slc:1 glc:1 end s_branch L_RESTORE_LDS L_RESTORE_VGPR_WAVE64: s_lshl_b32 s_restore_buf_rsrc2, s_restore_alloc_size, 8 //NUM_RECORDS in bytes (64 threads*4) if (SWIZZLE_EN) s_add_u32 s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_restore_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_mov_b32 s_restore_mem_offset_save, s_restore_mem_offset // restore start with v1, v0 will be the last s_add_u32 s_restore_mem_offset, s_restore_mem_offset, 256 s_mov_b32 m0, 1 //VGPR initial index value = 1 L_RESTORE_VGPR_WAVE64_LOOP: if(USE_MTBUF_INSTEAD_OF_MUBUF) tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 else buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 end s_waitcnt vmcnt(0) //ensure data ready v_movreld_b32 v0, v0 //v[0+m0] = v0 s_add_u32 m0, m0, 1 //next vgpr index s_add_u32 s_restore_mem_offset, s_restore_mem_offset, 256 //every buffer_load_dword does 256 bytes s_cmp_lt_u32 m0, s_restore_alloc_size //scc = (m0 < s_restore_alloc_size) ? 1 : 0 s_cbranch_scc1 L_RESTORE_VGPR_WAVE64_LOOP //VGPR restore (except v0) is complete? //s_set_gpr_idx_off // //Below part will be the restore shared vgpr part (new for gfx10) s_getreg_b32 s_restore_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_VGPR_SHARED_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_VGPR_SHARED_SIZE_SIZE) //shared_vgpr_size s_and_b32 s_restore_alloc_size, s_restore_alloc_size, 0xFFFFFFFF //shared_vgpr_size is zero? s_cbranch_scc0 L_RESTORE_V0 //no shared_vgpr used? jump to L_SAVE_LDS s_lshl_b32 s_restore_alloc_size, s_restore_alloc_size, 3 //Number of SHARED_VGPRs = shared_vgpr_size * 8 (non-zero value) //m0 now has the value of normal vgpr count, just add the m0 with shared_vgpr count to get the total count. //restore shared_vgpr will start from the index of m0 s_add_u32 s_restore_alloc_size, s_restore_alloc_size, m0 s_mov_b32 exec_lo, 0xFFFFFFFF s_mov_b32 exec_hi, 0x00000000 L_RESTORE_SHARED_VGPR_WAVE64_LOOP: buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 s_waitcnt vmcnt(0) //ensure data ready v_movreld_b32 v0, v0 //v[0+m0] = v0 s_add_u32 m0, m0, 1 //next vgpr index s_add_u32 s_restore_mem_offset, s_restore_mem_offset, 128 //every buffer_load_dword does 256 bytes s_cmp_lt_u32 m0, s_restore_alloc_size //scc = (m0 < s_restore_alloc_size) ? 1 : 0 s_cbranch_scc1 L_RESTORE_SHARED_VGPR_WAVE64_LOOP //VGPR restore (except v0) is complete? s_mov_b32 exec_hi, 0xFFFFFFFF //restore back exec_hi before restoring V0!! /* VGPR restore on v0 */ L_RESTORE_V0: if(USE_MTBUF_INSTEAD_OF_MUBUF) tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 else buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save slc:1 glc:1 end /* restore LDS */ ////////////////////////////// L_RESTORE_LDS: //Only need to check the first wave /* the first wave in the threadgroup */ s_and_b32 s_restore_tmp, s_restore_size, S_RESTORE_SPI_INIT_FIRST_WAVE_MASK s_cbranch_scc0 L_RESTORE_SGPR s_mov_b32 exec_lo, 0xFFFFFFFF //need every thread from now on //be consistent with SAVE although can be moved ahead s_and_b32 m0, s_restore_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_ENABLE_RESTORE_LDS_EXEC_HI s_mov_b32 exec_hi, 0x00000000 s_branch L_RESTORE_LDS_NORMAL L_ENABLE_RESTORE_LDS_EXEC_HI: s_mov_b32 exec_hi, 0xFFFFFFFF L_RESTORE_LDS_NORMAL: s_getreg_b32 s_restore_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE) //lds_size s_and_b32 s_restore_alloc_size, s_restore_alloc_size, 0xFFFFFFFF //lds_size is zero? s_cbranch_scc0 L_RESTORE_SGPR //no lds used? jump to L_RESTORE_VGPR s_lshl_b32 s_restore_alloc_size, s_restore_alloc_size, 6 //LDS size in dwords = lds_size * 64dw s_lshl_b32 s_restore_alloc_size, s_restore_alloc_size, 2 //LDS size in bytes s_mov_b32 s_restore_buf_rsrc2, s_restore_alloc_size //NUM_RECORDS in bytes if (SWIZZLE_EN) s_add_u32 s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_restore_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_and_b32 m0, s_wave_size, 1 s_cmp_eq_u32 m0, 1 s_mov_b32 m0, 0x0 s_cbranch_scc1 L_RESTORE_LDS_LOOP_W64 L_RESTORE_LDS_LOOP_W32: if (SAVE_LDS) buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset lds:1 s_waitcnt 0 end s_add_u32 m0, m0, 128 //every buffer_load_dword does 256 bytes s_add_u32 s_restore_mem_offset, s_restore_mem_offset, 128 //mem offset increased by 256 bytes s_cmp_lt_u32 m0, s_restore_alloc_size //scc=(m0 < s_restore_alloc_size) ? 1 : 0 s_cbranch_scc1 L_RESTORE_LDS_LOOP_W32 //LDS restore is complete? s_branch L_RESTORE_SGPR L_RESTORE_LDS_LOOP_W64: if (SAVE_LDS) buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset lds:1 s_waitcnt 0 end s_add_u32 m0, m0, 256 //every buffer_load_dword does 256 bytes s_add_u32 s_restore_mem_offset, s_restore_mem_offset, 256 //mem offset increased by 256 bytes s_cmp_lt_u32 m0, s_restore_alloc_size //scc=(m0 < s_restore_alloc_size) ? 1 : 0 s_cbranch_scc1 L_RESTORE_LDS_LOOP_W64 //LDS restore is complete? /* restore SGPRs */ ////////////////////////////// //s_getreg_b32 s_restore_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE) //spgr_size //s_add_u32 s_restore_alloc_size, s_restore_alloc_size, 1 //s_lshl_b32 s_restore_alloc_size, s_restore_alloc_size, 4 //Number of SGPRs = (sgpr_size + 1) * 16 (non-zero value) //s_lshl_b32 s_restore_alloc_size, s_restore_alloc_size, 3 //Number of SGPRs = (sgpr_size + 1) * 8 (non-zero value) L_RESTORE_SGPR: //need to look at it is wave32 or wave64 s_and_b32 m0, s_restore_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_RESTORE_SGPR_VMEM_WAVE64 if (SGPR_SAVE_USE_SQC) s_lshl_b32 s_restore_buf_rsrc2, s_sgpr_save_num, 2 //NUM_RECORDS in bytes else s_lshl_b32 s_restore_buf_rsrc2, s_sgpr_save_num, 7 //NUM_RECORDS in bytes (32 threads) end s_branch L_RESTORE_SGPR_CONT L_RESTORE_SGPR_VMEM_WAVE64: if (SGPR_SAVE_USE_SQC) s_lshl_b32 s_restore_buf_rsrc2, s_sgpr_save_num, 2 //NUM_RECORDS in bytes else s_lshl_b32 s_restore_buf_rsrc2, s_sgpr_save_num, 8 //NUM_RECORDS in bytes (64 threads) end L_RESTORE_SGPR_CONT: if (SWIZZLE_EN) s_add_u32 s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_restore_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_and_b32 m0, s_restore_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_RESTORE_SGPR_WAVE64 read_sgpr_from_mem_wave32(s_restore_tmp, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //save s0 to s_restore_tmp s_mov_b32 m0, 0x1 L_RESTORE_SGPR_LOOP_WAVE32: read_sgpr_from_mem_wave32(s0, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //PV: further performance improvement can be made s_waitcnt lgkmcnt(0) //ensure data ready s_movreld_b32 s0, s0 //s[0+m0] = s0 s_nop 0 // hazard SALU M0=> S_MOVREL s_add_u32 m0, m0, 1 //next sgpr index s_cmp_lt_u32 m0, s_sgpr_save_num //scc = (m0 < s_restore_alloc_size) ? 1 : 0 s_cbranch_scc1 L_RESTORE_SGPR_LOOP_WAVE32 //SGPR restore (except s0) is complete? s_mov_b32 s0, s_restore_tmp /* SGPR restore on s0 */ s_branch L_RESTORE_HWREG L_RESTORE_SGPR_WAVE64: read_sgpr_from_mem_wave64(s_restore_tmp, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //save s0 to s_restore_tmp s_mov_b32 m0, 0x1 //SGPR initial index value =1 //go on with with s1 L_RESTORE_SGPR_LOOP_WAVE64: read_sgpr_from_mem_wave64(s0, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //PV: further performance improvement can be made s_waitcnt lgkmcnt(0) //ensure data ready s_movreld_b32 s0, s0 //s[0+m0] = s0 s_nop 0 // hazard SALU M0=> S_MOVREL s_add_u32 m0, m0, 1 //next sgpr index s_cmp_lt_u32 m0, s_sgpr_save_num //scc = (m0 < s_restore_alloc_size) ? 1 : 0 s_cbranch_scc1 L_RESTORE_SGPR_LOOP_WAVE64 //SGPR restore (except s0) is complete? s_mov_b32 s0, s_restore_tmp /* SGPR restore on s0 */ /* restore HW registers */ ////////////////////////////// L_RESTORE_HWREG: s_mov_b32 s_restore_buf_rsrc2, 0x4 //NUM_RECORDS in bytes if (SWIZZLE_EN) s_add_u32 s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_restore_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_and_b32 m0, s_restore_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_RESTORE_HWREG_WAVE64 read_sgpr_from_mem_wave32(s_restore_m0, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //M0 read_sgpr_from_mem_wave32(s_restore_pc_lo, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //PC read_sgpr_from_mem_wave32(s_restore_pc_hi, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) read_sgpr_from_mem_wave32(s_restore_exec_lo, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //EXEC read_sgpr_from_mem_wave32(s_restore_exec_hi, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) read_sgpr_from_mem_wave32(s_restore_status, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //STATUS read_sgpr_from_mem_wave32(s_restore_trapsts, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //TRAPSTS //read_sgpr_from_mem_wave32(xnack_mask_lo, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //XNACK_MASK_LO //read_sgpr_from_mem_wave32(xnack_mask_hi, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //XNACK_MASK_HI read_sgpr_from_mem_wave32(s_restore_xnack_mask, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //XNACK_MASK read_sgpr_from_mem_wave32(s_restore_mode, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //MODE if(SAVE_RESTORE_HWID_DDID) read_sgpr_from_mem_wave32(s_restore_hwid1, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //HW_ID1 end s_branch L_RESTORE_HWREG_FINISH L_RESTORE_HWREG_WAVE64: read_sgpr_from_mem_wave64(s_restore_m0, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //M0 read_sgpr_from_mem_wave64(s_restore_pc_lo, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //PC read_sgpr_from_mem_wave64(s_restore_pc_hi, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) read_sgpr_from_mem_wave64(s_restore_exec_lo, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //EXEC read_sgpr_from_mem_wave64(s_restore_exec_hi, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) read_sgpr_from_mem_wave64(s_restore_status, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //STATUS read_sgpr_from_mem_wave64(s_restore_trapsts, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //TRAPSTS //read_sgpr_from_mem_wave64(xnack_mask_lo, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //XNACK_MASK_LO //read_sgpr_from_mem_wave64(xnack_mask_hi, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //XNACK_MASK_HI read_sgpr_from_mem_wave64(s_restore_xnack_mask, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //XNACK_MASK read_sgpr_from_mem_wave64(s_restore_mode, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //MODE if(SAVE_RESTORE_HWID_DDID) read_sgpr_from_mem_wave64(s_restore_hwid1, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) //HW_ID1 end L_RESTORE_HWREG_FINISH: s_waitcnt lgkmcnt(0) //from now on, it is safe to restore STATUS and IB_STS if(SAVE_RESTORE_HWID_DDID) L_RESTORE_DDID: s_mov_b32 m0, s_restore_hwid1 //virture ttrace support: The save-context handler records the SE/SA/WGP/SIMD/wave of the original wave s_ttracedata //and then can output it as SHADER_DATA to ttrace on restore to provide a correlation across the save-restore s_mov_b32 s_restore_buf_rsrc2, 0x4 //NUM_RECORDS in bytes if (SWIZZLE_EN) s_add_u32 s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0 //FIXME need to use swizzle to enable bounds checking? else s_mov_b32 s_restore_buf_rsrc2, 0x1000000 //NUM_RECORDS in bytes end s_and_b32 m0, s_restore_size, 1 s_cmp_eq_u32 m0, 1 s_cbranch_scc1 L_RESTORE_DDID_WAVE64 read_sgpr_from_mem_wave32(s_restore_ddid, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) s_branch L_RESTORE_DDID_FINISH L_RESTORE_DDID_WAVE64: read_sgpr_from_mem_wave64(s_restore_ddid, s_restore_buf_rsrc0, s_restore_mem_offset, SGPR_SAVE_USE_SQC) L_RESTORE_DDID_FINISH: s_waitcnt lgkmcnt(0) //s_mov_b32 m0, s_restore_ddid //s_ttracedata if (RESTORE_DDID_IN_SGPR18) s_mov_b32 s18, s_restore_ddid end end s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff //pc[47:32] //Do it here in order not to affect STATUS //for normal save & restore, the saved PC points to the next inst to execute, no adjustment needs to be made, otherwise: if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL)) s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 8 //pc[31:0]+8 //two back-to-back s_trap are used (first for save and second for restore) s_addc_u32 s_restore_pc_hi, s_restore_pc_hi, 0x0 //carry bit over end if ((EMU_RUN_HACK) && (EMU_RUN_HACK_RESTORE_NORMAL)) s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 4 //pc[31:0]+4 // save is hack through s_trap but restore is normal s_addc_u32 s_restore_pc_hi, s_restore_pc_hi, 0x0 //carry bit over end s_mov_b32 m0, s_restore_m0 s_mov_b32 exec_lo, s_restore_exec_lo s_mov_b32 exec_hi, s_restore_exec_hi s_and_b32 s_restore_m0, SQ_WAVE_TRAPSTS_PRE_SAVECTX_MASK, s_restore_trapsts s_setreg_b32 hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_PRE_SAVECTX_SHIFT, SQ_WAVE_TRAPSTS_PRE_SAVECTX_SIZE), s_restore_m0 s_setreg_b32 hwreg(HW_REG_SHADER_XNACK_MASK), s_restore_xnack_mask //restore xnack_mask s_and_b32 s_restore_m0, SQ_WAVE_TRAPSTS_POST_SAVECTX_MASK, s_restore_trapsts s_lshr_b32 s_restore_m0, s_restore_m0, SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT s_setreg_b32 hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT, SQ_WAVE_TRAPSTS_POST_SAVECTX_SIZE), s_restore_m0 //s_setreg_b32 hwreg(HW_REG_TRAPSTS), s_restore_trapsts //don't overwrite SAVECTX bit as it may be set through external SAVECTX during restore s_setreg_b32 hwreg(HW_REG_MODE), s_restore_mode //reuse s_restore_m0 as a temp register s_and_b32 s_restore_m0, s_restore_pc_hi, S_SAVE_PC_HI_RCNT_MASK s_lshr_b32 s_restore_m0, s_restore_m0, S_SAVE_PC_HI_RCNT_SHIFT s_lshl_b32 s_restore_m0, s_restore_m0, SQ_WAVE_IB_STS_RCNT_SHIFT s_mov_b32 s_restore_tmp, 0x0 //IB_STS is zero s_or_b32 s_restore_tmp, s_restore_tmp, s_restore_m0 s_and_b32 s_restore_m0, s_restore_pc_hi, S_SAVE_PC_HI_FIRST_REPLAY_MASK s_lshr_b32 s_restore_m0, s_restore_m0, S_SAVE_PC_HI_FIRST_REPLAY_SHIFT s_lshl_b32 s_restore_m0, s_restore_m0, SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT s_or_b32 s_restore_tmp, s_restore_tmp, s_restore_m0 s_and_b32 s_restore_m0, s_restore_status, SQ_WAVE_STATUS_INST_ATC_MASK s_lshr_b32 s_restore_m0, s_restore_m0, SQ_WAVE_STATUS_INST_ATC_SHIFT s_setreg_b32 hwreg(HW_REG_IB_STS), s_restore_tmp s_setreg_b32 hwreg(HW_REG_STATUS), s_restore_status s_barrier //barrier to ensure the readiness of LDS before access attemps from any other wave in the same TG //FIXME not performance-optimal at this time // s_rfe_b64 s_restore_pc_lo //Return to the main shader program and resume execution s_rfe_b64 s_restore_pc_lo // s_restore_m0[0] is used to set STATUS.inst_atc /**************************************************************************/ /* the END */ /**************************************************************************/ L_END_PGM: s_endpgm end /**************************************************************************/ /* the helper functions */ /**************************************************************************/ function write_sgpr_to_mem_wave32(s, s_rsrc, s_mem_offset, use_sqc, use_mtbuf) if (use_sqc) s_mov_b32 exec_lo, m0 //assuming exec_lo is not needed anymore from this point on s_mov_b32 m0, s_mem_offset s_buffer_store_dword s, s_rsrc, m0 glc:1 s_add_u32 s_mem_offset, s_mem_offset, 4 s_mov_b32 m0, exec_lo elsif (use_mtbuf) v_mov_b32 v0, s tbuffer_store_format_x v0, v0, s_rsrc, s_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 s_add_u32 s_mem_offset, s_mem_offset, 128 else v_mov_b32 v0, s buffer_store_dword v0, v0, s_rsrc, s_mem_offset slc:1 glc:1 s_add_u32 s_mem_offset, s_mem_offset, 128 end end function write_sgpr_to_mem_wave64(s, s_rsrc, s_mem_offset, use_sqc, use_mtbuf) if (use_sqc) s_mov_b32 exec_lo, m0 //assuming exec_lo is not needed anymore from this point on s_mov_b32 m0, s_mem_offset s_buffer_store_dword s, s_rsrc, m0 glc:1 s_add_u32 s_mem_offset, s_mem_offset, 4 s_mov_b32 m0, exec_lo elsif (use_mtbuf) v_mov_b32 v0, s tbuffer_store_format_x v0, v0, s_rsrc, s_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1 s_add_u32 s_mem_offset, s_mem_offset, 256 else v_mov_b32 v0, s buffer_store_dword v0, v0, s_rsrc, s_mem_offset slc:1 glc:1 s_add_u32 s_mem_offset, s_mem_offset, 256 end end function read_sgpr_from_mem_wave32(s, s_rsrc, s_mem_offset, use_sqc) s_buffer_load_dword s, s_rsrc, s_mem_offset glc:1 if (use_sqc) s_add_u32 s_mem_offset, s_mem_offset, 4 else s_add_u32 s_mem_offset, s_mem_offset, 128 end end function read_sgpr_from_mem_wave64(s, s_rsrc, s_mem_offset, use_sqc) s_buffer_load_dword s, s_rsrc, s_mem_offset glc:1 if (use_sqc) s_add_u32 s_mem_offset, s_mem_offset, 4 else s_add_u32 s_mem_offset, s_mem_offset, 256 end end