`
`some gate changes from c2
`
`Change 226410 on 2005/06/28 by smoss@smoss xenos_linuxorl
`
`testbench clean-up
`
`Change 226128 on 2005/06/24 by neelimam@neelimam_xenos_linux_orl
`
`<l. Changed path for the technology libraries to generic in tbh_gates.f for PA, HZ, VC
`and CP and in tbpd.f for SC.
`2. Updated README files for SC, VC and BIF.
`
`3. Added technology library ati_90n.v to tbgates spi.f.
`4. Added +tdefine+UNIT
`
`Change 224808 on 2005/06/13 by smoss@smoss_xenos_linuxorl
`
`updates
`
`Change 223796 on 2005/06/01 by smoss@smoss_xenos_linux_orl
`
`also corrected
`
`Change 223792 on 2005/06/01 by smoss@smoss_xenos_linuxorl
`
`corrected
`
`Change 222642 on 2005/05/20 by rramsey@rramseyxenos3_linux_orl
`
`get rid of extra include
`
`Change 222616 on 2005/05/20 by rramsey@rramseyxenos3_linuxorl
`
`Add headers for synth
`
`Change 222529 on 2005/05/19 by neelimam@neelimam_xenos_linux_orl
`
`<edited .f files to point to the correct netlist. Edited tbtrk_vc_out.v to remove the
`#4 delay in the clk.>
`
`Change 221262 on 2005/05/08 by smoss@smoss_ crayolalinuxorlregress
`
`update for sx gates
`
`Change 221171 on 2005/05/06 by smoss@smoss_crayola_linux_orl_regress
`
`1.
`
`remove gates from tb.f
`
`AMD1044_0237691
`
`ATI Ex. 2072
`IPR2023-00922
`Page 1 of 28
`
`
`
`2.
`
`3.
`
`update tbhgates
`
`update tb_sqsp.v for spi
`
`Change 221071 on 2005/05/06 by neelimam@neelimam_xenos_linux_orl
`
`<updated README files for gate simulations>
`
`Change 220567 on 2005/05/03 by neelimam@neelimam_xenos_linux_orl
`
`Edited files to point to the right netlist for A22
`
`Change 219797 on 2005/04/27 by rramsey@rramseyxenos3linuxorl
`
`submit SQECO6 changes to fix a performance issue in pixctl
`
`Change 219761 on 2005/04/27 by smoss@smoss_xenos_linux_orl
`
`remove gate references
`
`Change 219409 on 2005/04/25 by neelimam@neelimam_xenos_linuxorl
`
`<Edited tbh_gates.f for blocks HZ, PA, SC, SQ, VC to point to the PD netlist>
`
`Change 216633 on 2005/04/07 by smoss@smoss_crayola_linux_orl_regress
`
`enable dynamic clocking by default
`
`Change 216453 on 2005/04/06 by smoss@smoss_xenos_linux_orl
`
`I did not pass releasepl_xenos.pl, but must submit because: <lame excuse inserted
`here>
`
`updates for gate sims
`
`Change 215790 on 2005/04/01 by smoss@smoss_ crayola_linux_orl_regress
`
`I did not pass release_plxenos.pl, but must submit because: <lame excuse inserted
`here>
`
`get testbenches working for dft_eco_20
`
`Change 215379 on 2005/03/31 by hartogs@hartogsxenos_ linuxorl
`
`Change to SQ so that the RBBM can send transactions on the bus to other clients while
`the SQ fifo is full.
`
`Prior to this change, all RBBM transactions went thru the skid FIFO. After this
`
`change, only the RBBM
`transactions that are decoded on th
`
`Change 208531 on 2005/02/15 by rramsey@rramseyxenos3linuxorl
`
`AMD1044_0237692
`
`ATI Ex. 2072
`IPR2023-00922
`Page 2 of 28
`
`
`
`Changes for SQ_ECO4 and SQ_ECO5
`
`Change 203527 on 2005/01/14 by viviana@viviana_xenos2_linux_orl
`
`Added the clock adjuster deserializer inputs.
`
`Change 202408 on 2005/01/07 by viviana@viviana_xenos2_linux_orl
`
`Added DFT ECO's to the RTL code.
`
`Change 200981 on 2004/12/21 by danh@danhxenos5linuxorl
`
`I did not pass releaseplxenos.pl, but must submit because:
`
`this is just a text file
`
`Change 200354 on 2004/12/15 by danh@danh_xenos2_linux_orl
`
`I did not pass releasepl_xenos.pl, but must submit because:
`related to gate level simulation.
`
`these files are only
`
`Change 200339 on 2004/12/15 by danh@danh_xenos2_linux_orl
`
`I did not pass release_pl_xenos.pl, but must submit because: This is only a text file,
`it is a README file that explains how to run tb_sqsp gate level simulations
`
`Change 200318 on 2004/12/15 by danh@danh_xenos2_linux_orl
`
`these files are only used
`I did not pass releasepl_xenos.pl, but must submit because:
`for gate level simulation. Initial release of tb_sqsp gate level ".f" files.
`
`Change 194627 on 2004/11/03 by viviana@viviana_xenos2_linux_orl
`
`I did not pass releasepl_xenos.pl, but must submit because:
`These files are the TetraMAX models for the RAMS.
`
`Change 192659 on 2004/10/21 by rramsey@rramsey_xenos3_linux_orl
`
`state load change for tb_sqsp to fix suppress eop04
`
`Change 190504 on 2004/10/06 by rramsey@rramseyxenos3_linuxorl
`
`Add SQ_ECO3 to fix gpr dynamic/pool size bugs. This ECO is not enabled
`since it will not be fixed.
`
`Change 188818 on 2004/09/22 by bhankins@bhankins_real_xenos_linux_orl
`
`1. sx_exportcontrol _common.v - add hook for future implemetation of
`
`SX_ECO3 for properly generating the sx busy signal. Note that this
`
`AMD1044_0237693
`
`ATI Ex. 2072
`IPR2023-00922
`Page 3 of 28
`
`
`
`new code is hidden inside an “ifdef SX_ECO3, which is not defined.
`
`2. modify tbh_sqsp.v and tbmod
`
`Change 183607 on 2004/08/12 by smoss@smoss_xenos_linuxorl
`
`I did not pass release_pl_xenos.pl, but must submit because:
`testbench files
`
`these are virage
`
`Change 183350 on 2004/08/10 by rramsey@rramsey_xenos3_linux_orl
`
`Add SQ_ECO2 implementation. Fixes a phase generation problem in the SQ when power
`management
`is enabled.
`
`Change 183248 on 2004/08/10 by llefebvr@llefebvr_xenos_linux_marlboro
`
`This change adds is the code fix for SQ_ECOl by tying the SQ_CONSTANT_SENT_SP_SIMD2
`wire. This change is currently DISABLED.
`
`Change 179788 on 2004/07/21 by rramsey@rramseyxenos3_linuxorl
`
`modify testbench flush/bad_pipe synchronization to fix uber_rand
`timeout problem
`
`Change 177928 on 2004/07/12 by rramsey@rramseyxenos3linuxorl
`
`Check in Laurent's fixes for RSP fetch return problems in the emulator.
`
`Add definition of SQECO0 and ifdef previous rtl change to use it.
`
`Change 177331 on 2004/07/08 by rramsey@rramseyxenos3_linuxorl
`
`I did not pass release_pl_xenos.pl, but must submit because:
`these testbenches are not used in the sanity run
`
`add sq virage testbenches
`
`Change 176392 on 2004/06/30 by rramsey@rramseyxenos3_linuxorl
`
`Fix bug that happens when we get a map_copy with a const_write pending
`due to a full phys memory
`
`Change 176308 on 2004/06/30 by llefebvr@llefebvrxenoslinuxmarlboro
`
`This is fixing the RSP expand_enable bit.
`
`Change 176084 on 2004/06/29 by donaldl@donaldl_xenos2_linux_orl
`
`Commented out ifdef GCTEST so rsp compares work on tb_sqsp testbench.
`
`AMD1044_0237694
`
`ATI Ex. 2072
`IPR2023-00922
`Page 4 of 28
`
`
`
`Change 175630 on 2004/06/25 by rramsey@rramseyxenos3_linux_orl
`
`filter out vs_fetch_done events so they don't go to the sx
`add a transaction count for sx-bc events
`
`Fix serialize in vtx shader
`
`Change 175531 on 2004/06/25 by llefebvr@llefebvr_xenos_linux_marlboro
`
`Fixing max number of inputst+ param gen bug and the NOCEXEC_OPT, NO_LOOP_EXIT bugs.
`
`Change 175005 on 2004/06/22 by danh@danhxenos2_linuxorl
`
`Fixed SX_GATES section (ul_SX_SQ_* outputs are now only driven by sx_gates)
`
`Change 174919 on 2004/06/22 by rramsey@rramsey_xenos3_linux_orl
`
`I did not pass releasepl_xenos.pl, but must submit because:
`
`these files are only used in tb_sqsp
`
`trying to fix sxbe event tracking issues
`
`Change 174567 on 2004/06/18 by rramsey@rramseyxenos3_linux_orl
`
`I did not pass releasepl_xenos.pl, but must submit because:
`not part
`
`this testbench is
`
`of the sanity run
`
`fix sx_be event count mismatch error
`
`Change 173554 on 2004/06/14 by donaldl@donaldl_xenos2_linux_orl
`
`1. Changed DEBUSSY_PATH to VERDI_ROOT in buildtbgate.
`2. Updated vgt
`random scripts to use VCS and changed link requirement
`for tbfiles.
`
`3. Updated sq to sp trackers to not compare the shader pipe defined by
`
`ROM_SIMD_SEL[1:0] and ROMPIPESEL[
`
`Change 172977 on 2004/06/10 by rramsey@rramseyxenos3_linuxorl
`
`fix tpsp drivers in tb_sqsp to use post-steered pixel masks
`simd id was off by a clock in the sqtp and sqvc trackers
`
`Change 170316 on 2004/05/26 by llefebvr@llefebvrxenoslinuxmarlboro
`
`This is fixing the problem with the reduced size reservation stations.
`
`Change 170313 on 2004/05/26 by llefebvr@llefebvr_xenos_linux_marlboro
`
`This is fixing the tb_sqsp for predicated texture fetches.
`
`AMD1044_0237695
`
`ATI Ex. 2072
`IPR2023-00922
`Page 5 of 28
`
`
`
`Change 169247 on 2004/05/21 by bhankins@bhankinsreal_xenoslinux_orl
`
`add support for fake_hd_ fusebox
`
`Change 168875 on 2004/05/19 by rramsey@rramsey_xenos3_linux_orl
`
`adding fake_hd_fusebox to sq
`
`Change 168316 on 2004/05/17 by rramsey@rramseyxenos3_linux_orl
`
`timing fix for waterfall changes
`
`Change 167911 on 2004/05/14 by rramsey@rramseyxenos3linuxorl
`
`Remove realtime reads from alu and tex const mems
`
`to fix gate level sims.
`
`Add reset for some pix_ctl signals to try and fix synthesis issue with gpr_alloc
`state machine reset.
`
`Change sq_rbbm_skidbuf so it uses ati_skid_fifo_ctrl to help performance.
`
`Change 167871 on 2004/05/13 by danh@danh_xenos2_linux_orl
`
`Changed SX I/O ports in the SX_GATES section, so the SX gate level simulations will
`work properly
`
`Change 167557 on 2004/05/12 by rramsey@rramsey_xenos3_linux_orl
`
`Change arbiter so valid bits are updated before waterfall determination.
`
`Add a debug dump to sqalu.
`Change scalar KILL ops in vsp so they override vector kill results.
`
`Change ais output, queue, and seq so waterfall logic uses updated valid bits.
`
`Change 167129 on 2004/05/11 by smoss@smoss_ crayola_linuxorl regress
`
`<Orlando Hardware Regression Results >
`
`Change 166829 on 2004/05/09 by mearl@mearl_xenos_linuxorl
`
`1} Added top level registers to/from SC_A and CP_R (didn't turn on
`register yet, waiting for Frank L change to skid in CP).
`
`2) Added top level registers to/from PAV and SQ.
`
`3) Added top level registers from SCA to SPI, SP an
`
`Change 166815 on 2004/05/08 by mmantor@mmantor_xenos_linux_test
`
`<added top level registers between cgm and cp_r, mc_cp_r, sq_cp_r and changed fifo skid
`depth in rbbm, mc, and sq>
`
`AMD1044_0237696
`
`ATI Ex. 2072
`IPR2023-00922
`Page 6 of 28
`
`
`
`Change 166794 on 2004/05/08 by rramsey@rramseyxenos3_linux_orl
`
`change control flow pred optimization for pixel threads so it checks
`
`all pixels of a quad if any pixels in the quad are valid.
`
`add missing signal to sensitivity list in thread_arb.
`
`Change 166790 on 2004/05/08 by donaldl@donaldl_xenos2_linux_orl
`
`Created a credit-debit counter when sending VGT_SQ data
`to sq_ve_skid_buf.v.
`
`Change 166483 on 2004/05/06 by dclifton@dcliftonr400
`
`Updated to latest configuration
`
`Change 166355 on 2004/05/06 by rramsey@rramsey_xenos3_linux_orl
`
`fix loop/bool readback (read strobe was one clk too early)
`
`Change 166191 on 2004/05/05 by danh@danh_xenos2_linux_orl
`
`Gate level simulation changes and synthesis constraint changes.
`
`Change 166177 on 2004/05/05 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`- fix for dynamic gpr alloc (moved logic to update pix_max_q on a vtx alloc to a new
`state)
`
`Change 166084 on 2004/05/05 by mearl@mearl xenos_linux_orl
`
`1) Fixed a bug in the RSP logic in SQSPSX testbench.
`
`2) Fixed SC debug test.
`
`Change 165874 on 2004/05/04 by vromaker@vromakeremul_xenos_linux_marlboro
`
`- fix for dynamic gpr allocation (wrap head pointer when max was moved back to equal
`it)
`
`- initialization of pix_size and vtx_size in tbh_sqsp when dynamic allocation is forced
`
`Change 165822 on 2004/05/04 by rramsey@rramseyxenos3linuxorl
`
`Fix some wiring in the thread buffers that tied outputs together.
`
`Change DEBUSSY_PATH to VERDI_ROOT for buildtb pli source: ves only.
`
`Change 165816 on 2004/05/04 by bhankins@bhankins_real_xenos_linux_orl
`
`remove mc?_disables
`
`AMD1044_0237697
`
`ATI Ex. 2072
`IPR2023-00922
`Page 7 of 28
`
`
`
`Change 165618 on 2004/05/02 by rramsey@rramseyxenos3_linux_orl
`
`modify sq to use rotate64 and findtrailing1 from Mark Sprague.
`fix a gate-level reset issue in ais.
`
`fix I0_ scan bus widths in tb_sqsp.
`fix hd sms connections in sx.
`
`Change 165267 on 2004/04/30 by mmantor@mmantor_xenos_linux_test
`
`<checked in missing file>
`
`Change 165239 on 2004/04/30 by smoss@smoss_ crayolalinuxorlregress
`
`<Orlando Hardware Regression Results >
`
`Change 165153 on 2004/04/29 by rramsey@rramsey_xenos3_linux_orl
`
`Get rid of old pa hd fusebox and put the new one in
`Fix some sms connections in the SO
`
`Change 165070 on 2004/04/29 by mmantor@mmantor_xenos_linux_test
`
`<reordered the rbbm read bus,
`
`removed rbbm repeaters from sq and vc,
`
`fixed scan port
`
`width and hd_bclk moved internal to sq>
`
`Change 165009 on 2004/04/29 by smoss@smoss_crayola_linux_orl_regress
`
`I did not pass releasepl_xenos.pl, but must submit because:
`testbench specific
`
`these files are local
`
`Change 164749 on 2004/04/28 by bhankins@bhankins_real_xenos_linux_orl
`
`Remove the SX_hd_bclk input and connect the clock to sclk
`
`Change 164735 on 2004/04/28 by viviana@viviana_xenos2_linux_orl
`
`Added a missing memory and removed parf_stp.v no longer needed.
`Fixed some sq connections to the HD memories.
`
`Fixed the IO scan signals for the fsb.
`
`Change 164481 on 2004/04/27 by rramsey@rramsey_xenos3_linux_orl
`
`add export tests to sqsp regression
`fix is readback
`
`Change 164452 on 2004/04/27 by bhankins@bhankins_real_xenos_linux_orl
`
`fix bug to prevent ea exports from altering the exportbuffer valid bits
`
`AMD1044_0237698
`
`ATI Ex. 2072
`IPR2023-00922
`Page 8 of 28
`
`
`
`fix bug to prevent ea and em exports from writing to alpha memory
`
`change IOSX_scan and SX_IO_scan signals to 12 bits
`rename some signals for clarity
`
`Change 164078 on 2004/04/24 by bhankins@bhankinsrealxenoslinuxorl
`
`Advance SQ_SX_sp_exp_id by a clock,
`
`then fan out within sx, for timing
`
`Change 164018 on 2004/04/23 by llefebvr@llefebvr_xenos_linux_marlboro
`
`This is fixing Predicated conditionnal calls.
`
`Change 163965 on 2004/04/23 by rramsey@rramsey_xenos3_linux_orl
`
`add arbpriority register selects to readback strobe logic
`
`Change 163933 on 2004/04/23 by bhankins@bhankinsreal_xenos_linuxorl
`
`Reorder quads generated for em4 mem exports
`
`Fix bug in pixel mask generation
`
`&WMH . Modify resets to memory systems
`
`Replace multiple param cache input registers with a single register
`
`Change 163872 on 2004/04/23 by rramsey@rramseyxenos3_linux_orl
`
`Fixing readback of booleans and loop with real mems.
`
`Fix readback of ps program_size by correcting its top level connection.
`
`Change 163774 on 2004/04/22 by rramsey@rramseyxenos3_linux_orl
`
`fix readback of texconst mem
`
`Change 163670 on 2004/04/22 by rramsey@rramseyxenos3_linuxorl
`
`make sure instr_seq only clamps to base/max when doing relative addressing.
`add risms_stp blocks to th.f to get rid of verdi errors.
`fix serialize problem at beginning of clause in fetch cfs tracker.
`
`I did not pass release_pl_xenos.pl, but mus
`
`Change 163431 on 2004/04/20 by rramsey@rramseyxenos3_linuxorl
`
`fanout isphase to help timing
`
`Change 163399 on 2004/04/20 by rramsey@rramseyxenos3_linux_orl
`
`Timing fixes in alu_instr_seq.
`Change instr store so we can do a write every three clks when not blocked.
`
`Update rst_sms connections for sq and in tbsqsp.
`
`AMD1044_0237699
`
`ATI Ex. 2072
`IPR2023-00922
`Page 9 of 28
`
`
`
`Add ability to randomize rbbm inputs in tb_sqsp.
`
`Remove pa_valid check from second half of tex
`
`Change 163114 on 2004/04/19 by danh@danh_xenos2_linux_orl
`
`removed .v files that are do not need to be compiled
`
`Change 163112 on 2004/04/19 by danh@danh_xenos2_linux_orl
`
`The SPI, SP, RSP TST_<block>_rf_star_p1500[5] is now driven by !srst
`
`Change 163044 on 2004/04/19 by rramsey@RRAMSEYP4 xenoswin
`
`take out extra int_idle=0 that was keeping tests from timing out
`
`Change 163027 on 2004/04/19 by rramsey@rramsey_xenos3_linux_orl
`
`pick off thread count load signals one clock earlier to match up with changes in the
`SPI
`
`Change 162947 on 2004/04/18 by danh@danh_xenos2_linux_orl
`
`<block>_rfsms_stp.rst_sms is now driven by !TST_SP_rf_star_p1500[5], made gate level
`simulation changes to tb_sqsp.v
`
`Change 162930 on 2004/04/18 by hartogs@hartogs_xenos_linux_orl
`
`Chris Gray's GPU Layer change. This check-in is a significant change. It adds a gpu
`
`layer between testchip and chip.
`
`Change 162893 on 2004/04/17 by llefebvr@llefebvr_xenos_linux_marlboro
`
`This is fixing a texture constant update problem. Also making the skid 5 entries deep.
`
`Change 162531 on 2004/04/15 by danh@danh_xenos2_linux_orl
`
`Added +nowarn+NONPRT
`
`Change 162457 on 2004/04/15 by viviana@viviana_xenos2_linux_orl
`
`Memory changed from 4096x96 to 4096x32.
`
`Change 162454 on 2004/04/15 by viviana@viviana_xenos2_linux_orl
`
`File no longer used.
`
`Change 162446 on 2004/04/15 by rramsey@rramseyxenos3_linux_orl
`
`AMD1044_0237700
`
`ATI Ex. 2072
`IPR2023-00922
`Page 10 of 28
`
`
`
`sq timing fixes
`
`standardize the shift drivers in tb_sqsp
`
`I did not pass releasepl_xenos.pl, but must submit because:
`of release
`
`tb_sqsp is not part
`
`Change 162219 on 2004/04/14 by mearl@mearl_xenos_linux_orl
`
`1) Put in timing fix for the sc_packer
`2) More RSP fixes in tp_sqsp testbench
`
`Change 162204 on 2004/04/14 by danh@danh_xenos2_linux_orl
`
`Added reset state of q_vgt_send to resolve synthesis and gate level simulation problem
`
`Change 162203 on 2004/04/14 by danh@danh_xenos2_linux_orl
`
`Changed pathnames of all shader_*.ql_sq_export_simd_sel0a to
`
`shader*.sq_export_simd_sel_ 0a.q so the same pathnames can be used for gate level
`simulations
`
`Change 162190 on 2004/04/14 by vhopkins@vhopkins_xenos2_linux_orl
`
`Made timing changes
`
`Change 162121 on 2004/04/13 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`- fixes for dynamic gpr allocation mode
`
`- pix_ctl: need to select the pix_max to use for wrapping at the time the interp gpr
`addr is driven
`
`- gpr_alloc: need an extra delay state in the state machine under some dynamic alloc
`circumstances
`
`Change 161072 on 2004/04/08 by rramsey@rramseyxenos3linuxorl
`
`move a mux to help timing
`
`Change 160820 on 2004/04/08 by rramsey@rramseyxenos3_linux_orl
`
`remove exec_sm_idle check from loop and call instr decode logic to help timing.
`It should not be needed now that we have a copy of the status for each stage.
`
`Duplicate some logic in the pix thread buff to reduce fanout.
`
`Change 160726 on 2004/04/07 by danh@danh_xenos2_linux_orl
`
`Fixed uncontrollable ATPG inputs and .syn wire_load issues
`I did not pass releasepl_xenos.pl, but must submit because: spi.syn (synthesis
`
`only file) was the only file that changed since I last ran releaseplxenos.pl
`
`AMD1044_0237701
`
`ATI Ex. 2072
`IPR2023-00922
`Page 11 of 28
`
`
`
`Change 160583 on 2004/04/07 by rramsey@rramseyxenos3_linux_orl
`
`reduce fanout on logical address to help timing
`
`Change 160526 on 2004/04/07 by mmantor@mmantor_xenos_linux_test
`
`<tied HZ repeater rbbm bus to no-connects to remove from chip port list
`
`sent const_sel signal to sp ane rsp a clock earlier and removed fanout on input
`made sp determine adjust phase to remove same problem of tst insertion >
`
`Change 160525 on 2004/04/07 by mmantor@mmantorxenos linuxtest
`
`<removed FPLLFSB clocks from chip.v and added skid buffering to VGT and SQ for RBBM
`bus repeaters>
`
`Change 160357 on 2004/04/06 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`status update
`
`Change 160308 on 2004/04/06 by rramsey@rramseyxenos3_linux_orl
`
`fixes a problem with the cf tracker
`
`Change 160269 on 2004/04/06 by rramsey@rramseyxenos3linuxorl
`
`changes to tb_sqsp to help get trackers working with dynamic gpr enabled
`
`Change 160144 on 2004/04/06 by rramsey@fl_rramsey1t_r400_win
`
`update with regression results
`
`Change 160127 on 2004/04/06 by mmantor@mmantorxenos linuxtest
`
`<change name of FSB signal name from FSB_SI_stop_clock to FSB_CLK_stop_clk and
`connected fsbclk to FPLL_FSB_clkout8
`
`added an output port temporily for FPLL_FSBclkout8 to drive testchip models.
`Added comments to chip.tree for
`SHIFT bus seria
`
`Change 159901 on 2004/04/05 by smoss@smoss_xenos_linux_orl
`
`I did not pass release_pl_xenos.pl, but must submit because:
`files and do not affect releasepl xenos
`
`these are local testbench
`
`Change 159803 on 2004/04/03 by rramsey@rramsey_xenos3_linux_orl
`
`simplify busy extender.
`
`AMD1044_0237702
`
`ATI Ex. 2072
`IPR2023-00922
`Page 12 of 28
`
`
`
`timing fixes for ctl_flow_seq.
`
`clean up sq clk_en and wake_up signals to other blocks.
`change back alu cf tracker to match emulator when skipping serial bits.
`
`both cf trackers need to look at a new cip signal.
`
`fix a bug in the
`
`Change 159796 on 2004/04/03 by mmantor@mmantorxenoslinux_test
`
`< Partitioned the RBBM Write Bus into 4 segments with three repeaters and
`
`all the necessary change for the full chip,
`benches
`
`tb_sqps, gc,
`
`tp4_tce and tc test
`
`Driver
`
`(CP_R or Repeater) => Recievers (all on common net)
`
`CP_R =>
`
`TCM, TCF
`
`Change 159476 on 2004/04/02 by rramsey@rramseyxenos3_linux_orl
`
`now that mike checked in the updated memory,
`
`this needs to be changed again
`
`This changelist fixes sanity failure: 159465
`
`Change 159474 on 2004/04/02 by mmantor@mmantorxenos linuxtest
`
`<fix regression by checking in the real file that I missed before. This changelist
`
`fixes sanity failure: <159459>>
`
`Change 159471 on 2004/04/02 by rramsey@rramseyxenos3linuxorl
`
`change port name back to rst_sms_a until mike can figure out what he needs to do
`This changelist fixes sanity failure: 159465
`
`Change 159459 on 2004/04/02 by mmantor@mmantorxenos_linuxtest
`
`<appears I missed a file checkin so here it is>
`
`Change 159415 on 2004/04/01 by mearl@mearlxenos_linuxorl
`
`1) Fixed bug in packer where deallocs were not being sent with
`
`the last quad of prim.
`
`2) Turned off RT trackers for output of packer and SC/SQ interface.
`
`3) Updated RSP in tb_sqsp testbench.
`
`Change 159186 on 2004/04/01 by rramsey@rramsey_xenos3_linux_orl
`
`yet another fix for random tests
`
`Change 159116 on 2004/03/31 by delifton@fl_dcliftonlt
`
`Updated with more bist changes.
`
`AMD1044_0237703
`
`ATI Ex. 2072
`IPR2023-00922
`Page 13 of 28
`
`
`
`Change 159082 on 2004/03/31 by smoss@smoss_xenos_chip
`
`changed port from rst_sms to rst_sms_a to match memory
`
`Change 159003 on 2004/03/31 by vromaker@vromaker_emul_xenos_linuxmarlboro
`
`status update
`
`Change 159002 on 2004/03/31 by llefebvr@llefebvr_laptop_xenos_emu
`
`status update
`
`Change 158983 on 2004/03/31 by mmantor@mmantor_xenos_linux_test
`
`<new virage memories and instruction store mems split into three columns>
`
`Change 158975 on 2004/03/31 by rramsey@rramseyxenos3_linuxorl
`
`increase thread fifo from 4
`
`to 8 entries
`
`Change 158969 on 2004/03/31 by rramsey@RRAMSEYP4_xenos_win
`
`update status
`
`Change 158869 on 2004/03/31 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`status update, picked test
`
`Change 158856 on 2004/03/31 by llefebvr@llefebvr_laptopxenos_emu
`
`update status
`
`Change 158855 on 2004/03/31 by llefebvr@llefebvr_xenos_linuxmarlboro
`
`Qualifying the restart of the texture constant store on the RTS of the RBBM.
`
`Change 158844 on 2004/03/31 by rramsey@RRAMSEYP4 xenos_win
`
`update with 3/30 status
`
`Change 158636 on 2004/03/30 by rramsey@rramsey_xenos3_linux_orl
`
`fix tp_rsp_simd connection to repeater regs.
`vtx_ctl was dropping events if they came in when the thread fifo was full.
`
`Change 158549 on 2004/03/30 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`status update, picked test
`
`AMD1044_0237704
`
`ATI Ex. 2072
`IPR2023-00922
`Page 14 of 28
`
`
`
`Change 158435 on 2004/03/29 by llefebvr@llefebvr_laptop_xenos_emu
`
`status update
`
`Change 158381 on 2004/03/29 by llefebvr@llefebvr_laptop_xenos_emu
`
`update status
`
`Change 158379 on 2004/03/29 by llefebvr@llefebvr_laptop_xenos_emu
`
`update status
`
`Change 158371 on 2004/03/29 by vromaker@vromakeremulxenos linuxmarlboro
`
`status update, new test picked
`
`I did not pass releasepl_xenos.pl, but must submit because: <lame excuse inserted
`here>
`
`Change 158359 on 2004/03/29 by rramsey@rramseyxenos3_linuxorl
`
`add sc_sq valid to int_idle check.
`
`change sq_sp int_idle check to look at instr_stall instead of instrstart.
`fix alu cfs tracker to ignore serial bits in pred instr when pred_dirty causes
`a clause boundary.
`
`Change 158309 on 2004/03/29 by rramsey@RRAMSEYP4 xenoswin
`
`update with 3/29 results
`
`Change 158294 on 2004/03/29 by mmantor@mmantor_xenos_linux_test
`
`<added the shift_block to the tc test bench, made a change for timing in the
`sq_alu_instr_seq.v and added an additional pulse on shiftblk_clkin to move data from
`output register of deserializer to the shift register and removed the ifdef synthesis c
`
`Change 158292 on 2004/03/29 by bhankins@bhankins_real_xenos_linuxorl
`
`add unused virage test outputs to sx
`
`Change 158222 on 2004/03/26 by rramsey@rramsey_xenos3_linux_orl
`
`Change fetch arb so tp fetch can go if ve queue has a mega, but the ve
`
`count is full so the mega can't go.
`
`Add a fifo between the vsr_load and gpr_load machines so we can keep the
`vsr loads going when we have event pushes pending.
`
`Change 158114 on 2004/03/26 by bhankins@bhankinsrealxenoslinuxorl
`
`AMD1044_0237705
`
`ATI Ex. 2072
`IPR2023-00922
`Page 15 of 28
`
`
`
`add components for HD mems
`
`Change 158080 on 2004/03/26 by bhankins@bhankinsreal_xenoslinuxorl
`
`Add newest changes to memory systems
`
`Change 158074 on 2004/03/26 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`- status update, picked a test
`
`Change 158069 on 2004/03/26 by mmantor@mmantorxenos linuxtest
`
`<some shift in port connections and aluconst memory freelist backpressure problem>
`
`Change 158053 on 2004/03/26 by llefebvr@llefebvr_xenos_emu_montreal
`
`status update
`
`Change 157921 on 2004/03/25 by llefebvr@llefebvr_xenos_emu_montreal
`
`status update
`
`Change 157892 on 2004/03/25 by rramsey@RRAMSEYP4 xenos_win
`
`update status
`
`Change 157891 on 2004/03/25 by rramsey@rramsey_xenos3_linux_orl
`
`enable flush logic
`
`Change 157890 on 2004/03/25 by rramsey@rramseyxenos3_linux_orl
`
`only updating generation timestamp so the whole rf system is consistent
`
`Change 157873 on 2004/03/25 by llefebvr@llefebvr_xenos_emu_montreal
`
`update status
`
`Change 157825 on 2004/03/25 by mearl@mearlxenos_linuxorl
`
`1) Updated buildkdb script for XENOS
`
`2) Updated the multiple rf_rm_enable signals to only one
`
`Change 157819 on 2004/03/25 by llefebvr@llefebvr_xenos_emu_montreal
`
`update status
`
`AMD1044_0237706
`
`ATI Ex. 2072
`IPR2023-00922
`Page 16 of 28
`
`
`
`Change 157809 on 2004/03/25 by rramsey@rramseyxenos3_linux_orl
`
`new rf memories from Vivian
`
`hook all rf mems
`
`to a common rmb enable
`
`Change 157766 on 2004/03/25 by rramsey@rramsey_xenos3_linux_orl
`
`fix a couple of Virage memory connections
`
`remove extra clk delay on spsx no_compare signals so they line up with sp outputs
`cfs tracker now ignores serial bits for unexecuted conditionals
`
`add rfsms files to tb.f to get rid of verdi load errors
`
`Change 157735 on 2004/03/24 by llefebvr@llefebvr_xenos_emu_montreal
`
`status update
`
`Change 157733 on 2004/03/24 by llefebvr@llefebvr_xenos_linux_marlboro
`
`The texture store was dropping a full constant update when staling.
`
`Change 157454 on 2004/03/23 by dclifton@fl_dcliftonlt
`
`Updated for latest ram and test blocks
`
`Change 157378 on 2004/03/23 by rramsey@RRAMSEYP4 xenos_win
`
`update status, pick new test
`
`Change 157375 on 2004/03/23 by rramsey@rramseyxenos3_linux_orl
`
`wait_forpixshaderdone logic was looking at vs_event_id instead of
`
`ps_event_id
`
`Change 157356 on 2004/03/23 by rramsey@rramseyxenos3linuxorl
`
`turn off debug prints
`
`Change 157344 on 2004/03/23 by rramsey@RRAMSEYP4_xenos_win
`
`pick a test
`
`Change 157333 on 2004/03/23 by danh@danhxenos2_linuxorl
`
`<block>_rfsms_p1500 and ati_deserializer instantiations
`
`Change 157330 on 2004/03/23 by rramsey@rramsey_xenos3_linux_orl
`
`new Virage mems and connections for deserializer
`
`AMD1044_0237707
`
`ATI Ex. 2072
`IPR2023-00922
`Page 17 of 28
`
`
`
`Fix internal cfs trackers and turn them back on in tb_sqsp
`
`Change 157272 on 2004/03/23 by rramsey@RRAMSEYP4 xenos_win
`
`update with 3/23 status
`
`Change 157177 on 2004/03/22 by mearl@mearl_xenos_linuxorl
`
`1) Added coverage pragmas to SC code
`
`2) Updated to new Virage Memory Systems
`
`Change 157135 on 2004/03/22 by llefebvr@llefebvrxenosemu_montreal
`
`updating status.
`
`Change 157132 on 2004/03/22 by llefebvr@llefebvr_xenos_linux_marlboro
`
`The rtr of the ALU store was kept high too long allowing 1 constant overwrite at the
`
`begining of state.
`
`Change 157029 on 2004/03/21 by viviana@viviana_xenos2_linux_orl
`
`Memory lists for the Virage memories.
`
`Change 156798 on 2004/03/19 by llefebvr@llefebvr_xenos_linuxmarlboro
`
`Fixing the MINI counter to use ve_ fifodepths11reqfifodepth instead of
`ve_fifo_depths_1l_ reqfifodepth +2.
`
`Change 156740 on 2004/03/19 by llefebvr@llefebvr_xenos_linux_marlboro
`
`This is fixing the read enable and invalid read addresses on the cfe rams.
`
`Change 156654 on 2004/03/19 by vromaker@vromakeremul_xenos_linuxmarlboro
`
`status update, picked test
`
`Change 156348 on 2004/03/18 by llefebvr@llefebvr_xenos_emu_montreal
`
`update status
`
`Change 156318 on 2004/03/18 by vromaker@vromakeremulxenos linuxmarlboro
`
`updated status, signed up to investigate flow control stall failure
`
`Change 156266 on 2004/03/18 by rramsey@RRAMSEYP4_ xenos_win
`
`picked a test
`
`AMD1044_0237708
`
`ATI Ex. 2072
`IPR2023-00922
`Page 18 of 28
`
`
`
`Change 156204 on 2004/03/18 by llefebvr@llefebvr_xenos_emu_montreal
`
`status update
`
`Change 156178 on 2004/03/18 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`updated status,
`
`took open tests
`
`Change 156169 on 2004/03/18 by llefebvr@llefebvr_xenos_emu_montreal
`
`updating status
`
`Change 156130 on 2004/03/18 by mmantor@mmantorxenos linuxtest
`
`<removed last pipe stage on sp outputs since we beleive not needed for top level
`connections>
`
`Change 155984 on 2004/03/17 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`status update
`
`Change 155943 on 2004/03/17 by llefebvr@llefebvr_xenos_emu_montreal
`
`updating status.
`
`Change 155936 on 2004/03/17 by rramsey@RRAMSEYP4 xenoswin
`
`xenos regress status
`
`Change 155776 on 2004/03/17 by llefebvr@llefebvr_xenos_linux_marlboro
`
`This is fixing the free list to stall when it is full.
`
`Change 155692 on 2004/03/16 by rramsey@rramseyxenos3_linux_orl
`
`Change cfs to send an update for end_of_shader if it ends on an empty clause. This
`keeps us from hanging due to the fact that nothing gets sent to the ais to trigger
`
`an ais done.
`Add clk adjust block to S5Q.
`
`Fix event pulses from vtx thread buffer so
`
`Change 155668 on 2004/03/16 by danh@danh_xenos2_linux_orl
`
`Added ati_clock_adjuster
`
`Change 155389 on 2004/03/16 by mmantor@mmantorxenos_linuxtest
`
`AMD1044_0237709
`
`ATI Ex. 2072
`IPR2023-00922
`Page 19 of 28
`
`
`
`<refresh on sq rf mem system, new sq hd mem system, sq internal tracker fixes for sq
`
`signal replication, sq internal cfs tracker disable for now, changed name of hd
`
`fusebox, enabled real memories all the time in tbhsqsp, added new top level register
`
`Change 154880 on 2004/03/12 by donaldl@donaldl_xenos2_ linuxorl
`
`Fanned out SQSP signals (from the SQ to the SPs and SPIs).
`
`Change 154755 on 2004/03/12 by mmantor@mmantor_xenos_linux_test
`
`<added sc_bc top level register module and removed an old memory>
`
`Change 154479 on 2004/03/11 by vromaker@vromaker_emul_xenos_linux_marlboro
`
`- connected the no_earlythread_termination bit of the flow_control register
`to the alu instr seq modules
`
`Change 154470 on 2004/03/11 by mearl@mearlxenos_linuxorl
`
`Added new memories to SC and HZ. Added RSP to tb_sqsp testbench.
`
`Change 154414 on 2004/03/11 by rramsey@rramseyxenos3_linux_orl
`
`ifdef some debug only display calls
`
`Change 154347 on 2004/03/11 by mmantor@mmantor_xenos_linux_test
`
`<removed old memories and added new real memories>
`
`Change 154218 on 2004/03/10 by mmantor@mmantorxenos_linuxtest
`
`<added rf memory star processor, memories, with some clean up>
`
`Change 154093 on 2004/03/10 by rramsey@rramseyxenos3linuxorl
`
`Get rid of verdi build errors for gc.
`
`Fix cfl eject logic in thread_arb- grouping for if statement was wrong
`so cfl was getting ejected incorrectly.
`
`Change eject in cfs so we are less likely to eject a thread if the other
`cfs becomes available before
`
`Change 154006 on 2004/03/10 by vromaker@vromakeremulxenos linuxmarlboro
`
`- changes to AIS update integrated from r400
`- fixed vtx shaders for two tests
`
`- commented out
`
`lod correct compare from sc out_compare tracker
`
`Change 153993 on 2004/03/10 by bhankins@bhankinsrealxenoslinuxorl
`
`AMD1044_0237710
`
`ATI Ex. 2072
`IPR2023-00922
`Page 20 of 28
`
`
`
`Split the SX_SP_vtx_data[0-5] signals from the SX into two sets of registered outputs.
`
`Change 153935 on 2004/03/10 by bhankins@bhankinsreal_xenoslinuxorl
`
`Add missing sx memtest outputs.
`
`Change 153766 on 2004/03/09 by llefebvr@llefebvr_xenos_linux_marlboro
`
`Put a reset on loop_rep_on signal. Fixes r400sq_loop_rep03/04.cpp
`
`Change 153701 on 2004/03/09 by bhankins@bhankins realxenos linuxorl
`
`Update the SX memory systems to 90nm.
`
`Change 153595 on 2004/03/09 by rramsey@rramsey_xenos3_linux_orl
`
`fix hang condition that could happen if contexts were loading the max number of
`
`constants and then selectively updating some. This fixes the hang for bug 3297.
`
`Change 153390 on 2004/03/08 by donaldl@donaldl_xenos2_linux_orl
`
`Qualified RSP comparing of data with sq_vec_fetch_type to fix erroneous mismatches.
`
`Change 153154 on 2004/03/06 by smoss@smoss xenos_chip
`
`coverage updates
`
`Change 152942 on 2004/03/05 by mmantor@mmantor_xenos_linux_test
`
`<Integrated aluconst mem fixes from r400, changes const memes in system_sq.vcpp,
`removed or disabled realtime register writes and fixed freelist