Changes between Version 32 and Version 33 of ClrxAsmGallium


Ignore:
Timestamp:
Feb 7, 2018, 9:00:40 PM (17 months ago)
Author:
trac
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • ClrxAsmGallium

    v32 v33  
    156156<h3>.dims</h3>
    157157<p>Syntax: .dims DIMENSIONS</p>
    158 <p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Defines what dimensions
     158<p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Define what dimensions
    159159(from list: x, y, z) will be used to determine space of the kernel execution.</p>
    160160<h3>.driver_version</h3>
     
    177177<h3>.floatmode</h3>
    178178<p>Syntax: .floatmode BYTE-VALUE</p>
    179 <p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Defines float-mode.
     179<p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Define float-mode.
    180180Set floatmode (FP_ROUND and FP_DENORM fields of the MODE register). Default value is 0xc0.</p>
    181181<h3>.gds_segment_size</h3>
     
    203203<p>Syntax: .hsa_dims DIMENSIONS</p>
    204204<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
    205 LLVM version is 4.0.0 or later. Defines what dimensions (from list: x, y, z) will be used
     205LLVM version is 4.0.0 or later. Define what dimensions (from list: x, y, z) will be used
    206206to determine space of the kernel execution in kernel HSA configuration.</p>
    207207<h3>.hsa_dx10clamp</h3>
     
    216216<p>Syntax: .hsa_floatmode BYTE-VALUE</p>
    217217<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
    218 LLVM version is 4.0.0 or later. Defines float-mode in kernel HSA configuration.
     218LLVM version is 4.0.0 or later. Define float-mode in kernel HSA configuration.
    219219Set floatmode (FP_ROUND and FP_DENORM fields of the MODE register). Default value is 0xc0.</p>
    220220<h3>.hsa_ieeemode</h3>
     
    225225<p>Syntax: .hsa_localsize SIZE</p>
    226226<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
    227 LLVM version is 4.0.0 or later. Defines initial local memory size used by kernel in
     227LLVM version is 4.0.0 or later. Define initial local memory size used by kernel in
    228228kernel HSA configuration.</p>
    229229<h3>.hsa_pgmrsrc1</h3>
    230230<p>Syntax: .hsa_pgmrsrc1 VALUE</p>
    231231<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
    232 LLVM version is 4.0.0 or later. Defines value of the PGMRSRC1 in kernel HSA configuration.</p>
     232LLVM version is 4.0.0 or later. Define value of the PGMRSRC1 in kernel HSA configuration.</p>
    233233<h3>.hsa_pgmrsrc2</h3>
    234234<p>Syntax: .hsa_pgmrsrc2 VALUE</p>
    235235<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
    236 LLVM version is 4.0.0 or later. Defines value of the PGMRSRC2 in kernel HSA configration.
     236LLVM version is 4.0.0 or later. Define value of the PGMRSRC2 in kernel HSA configration.
    237237If dimensions is set then bits that controls dimension setup will be ignored.
    238238SCRATCH_EN bit will be ignored.</p>
     
    240240<p>Syntax: .hsa_priority PRIORITY</p>
    241241<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
    242 LLVM version is 4.0.0 or later. Defines priority (0-3) in kernel HSA configuration.</p>
     242LLVM version is 4.0.0 or later. Define priority (0-3) in kernel HSA configuration.</p>
    243243<h3>.hsa_privmode</h3>
    244244<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
     
    248248<p>Syntax: .hsa_scratchbuffer SIZE</p>
    249249<p>This pseudo-op must be inside kernel configuration (<code>.config</code>) and can ben used only if
    250 LLVM version is 4.0.0 or later. Defines scratchbuffer size in kernel HSA configuration.</p>
     250LLVM version is 4.0.0 or later. Define scratchbuffer size in kernel HSA configuration.</p>
    251251<h3>.hsa_sgprsnum</h3>
    252252<p>Syntax: .hsa_sgprsnum REGNUM</p>
     
    319319<h3>.localsize</h3>
    320320<p>Syntax: .localsize SIZE</p>
    321 <p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Defines initial
     321<p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Define initial
    322322local memory size used by kernel.</p>
    323323<h3>.machine</h3>
     
    333333<p>Syntax: .pgmrsrc1 VALUE</p>
    334334<p>This pseudo-op must be inside kernel configuration (<code>.config</code>).
    335 Defines value of the PGMRSRC1.</p>
     335Define value of the PGMRSRC1.</p>
    336336<h3>.pgmrsrc2</h3>
    337337<p>Syntax: .pgmrsrc2 VALUE</p>
    338338<p>This pseudo-op must be inside kernel configuration (<code>.config</code>).
    339 Defines value of the PGMRSRC2. If dimensions is set then bits that controls dimension setup
     339Define value of the PGMRSRC2. If dimensions is set then bits that controls dimension setup
    340340will be ignored. SCRATCH_EN bit will be ignored.</p>
    341341<h3>.priority</h3>
    342342<p>Syntax: .priority PRIORITY</p>
    343 <p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Defines priority (0-3).</p>
     343<p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Define priority (0-3).</p>
    344344<h3>.private_elem_size</h3>
    345345<p>Syntax: .private_elem_size ELEMSIZE</p>
     
    378378<h3>.scratchbuffer</h3>
    379379<p>Syntax: .scratchbuffer SIZE</p>
    380 <p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Defines scratchbuffer size.</p>
     380<p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Define scratchbuffer size.</p>
    381381<h3>.scratchsym</h3>
    382382<p>Syntax: .scratchsym SYMBOL</p>
     
    389389It counts SGPR registers including VCC, FLAT_SCRATCH and XNACK_MASK.</p>
    390390<h3>.spilledsgprs</h3>
    391 <p>Syntax: .sgpilledsgprs REGNUM</p>
     391<p>Syntax: .spilledsgprs REGNUM</p>
    392392<p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Set number of scalar
    393393registers to spill in scratch buffer. It have meaning for LLVM 3.9 or later.</p>
    394394<h3>.spilledvgprs</h3>
    395 <p>Syntax: .sgpilledvgprs REGNUM</p>
     395<p>Syntax: .spilledvgprs REGNUM</p>
    396396<p>This pseudo-op must be inside kernel configuration (<code>.config</code>). Set number of vector
    397397registers to spill in scratch buffer. It have meaning for LLVM 3.9 or later. </p>