| /dports/lang/rust/rustc-1.58.1-src/src/llvm-project/llvm/docs/ |
| H A D | Bugpoint.rst | 12 can be used to debug three types of failures: optimizer crashes, miscompilations
17 the file down to a small example which triggers the crash.
33 where each test of the program (which requires executing it) takes a long time.
65 If an optimizer or code generator crashes, ``bugpoint`` will try as hard as it
66 can to reduce the list of passes (for optimizer crashes) and the size of the
67 test program. First, ``bugpoint`` figures out which combination of optimizer
107 which are causing the miscompilation, then reduce the portion of the test
108 program which is being miscompiled. The miscompilation debugger assumes that
160 * ``bugpoint`` can produce IR which contains long names. Run ``opt
175 Turn on ``-debug-only=instcombine`` and see which transformations within
[all …]
|
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/llvm-devel/llvm-project-f05c95f10fc1d8171071735af8ad3a9e87633120/llvm/docs/ |
| H A D | Bugpoint.rst | 12 can be used to debug three types of failures: optimizer crashes, miscompilations
17 the file down to a small example which triggers the crash.
33 where each test of the program (which requires executing it) takes a long time.
65 If an optimizer or code generator crashes, ``bugpoint`` will try as hard as it
66 can to reduce the list of passes (for optimizer crashes) and the size of the
67 test program. First, ``bugpoint`` figures out which combination of optimizer
107 which are causing the miscompilation, then reduce the portion of the test
108 program which is being miscompiled. The miscompilation debugger assumes that
160 * ``bugpoint`` can produce IR which contains long names. Run ``opt
175 Turn on ``-debug-only=instcombine`` and see which transformations within
[all …]
|
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/wasi-compiler-rt13/llvm-project-13.0.1.src/llvm/docs/ |
| H A D | Bugpoint.rst | 12 can be used to debug three types of failures: optimizer crashes, miscompilations
17 the file down to a small example which triggers the crash.
33 where each test of the program (which requires executing it) takes a long time.
65 If an optimizer or code generator crashes, ``bugpoint`` will try as hard as it
66 can to reduce the list of passes (for optimizer crashes) and the size of the
67 test program. First, ``bugpoint`` figures out which combination of optimizer
107 which are causing the miscompilation, then reduce the portion of the test
108 program which is being miscompiled. The miscompilation debugger assumes that
160 * ``bugpoint`` can produce IR which contains long names. Run ``opt
175 Turn on ``-debug-only=instcombine`` and see which transformations within
[all …]
|
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/wasi-compiler-rt12/llvm-project-12.0.1.src/llvm/docs/ |
| H A D | Bugpoint.rst | 12 can be used to debug three types of failures: optimizer crashes, miscompilations
17 the file down to a small example which triggers the crash.
33 where each test of the program (which requires executing it) takes a long time.
65 If an optimizer or code generator crashes, ``bugpoint`` will try as hard as it
66 can to reduce the list of passes (for optimizer crashes) and the size of the
67 test program. First, ``bugpoint`` figures out which combination of optimizer
107 which are causing the miscompilation, then reduce the portion of the test
108 program which is being miscompiled. The miscompilation debugger assumes that
160 * ``bugpoint`` can produce IR which contains long names. Run ``opt
175 Turn on ``-debug-only=instcombine`` and see which transformations within
[all …]
|
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/tinygo/tinygo-0.14.1/llvm-project/llvm/docs/ |
| H A D | Bugpoint.rst | 12 can be used to debug three types of failures: optimizer crashes, miscompilations
17 the file down to a small example which triggers the crash.
33 where each test of the program (which requires executing it) takes a long time.
65 If an optimizer or code generator crashes, ``bugpoint`` will try as hard as it
66 can to reduce the list of passes (for optimizer crashes) and the size of the
67 test program. First, ``bugpoint`` figures out which combination of optimizer
107 which are causing the miscompilation, then reduce the portion of the test
108 program which is being miscompiled. The miscompilation debugger assumes that
160 * ``bugpoint`` can produce IR which contains long names. Run ``opt
175 Turn on ``-debug-only=instcombine`` and see which transformations within
[all …]
|
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/llvm13/llvm-project-13.0.1.src/llvm/docs/ |
| H A D | Bugpoint.rst | 12 can be used to debug three types of failures: optimizer crashes, miscompilations
17 the file down to a small example which triggers the crash.
33 where each test of the program (which requires executing it) takes a long time.
65 If an optimizer or code generator crashes, ``bugpoint`` will try as hard as it
66 can to reduce the list of passes (for optimizer crashes) and the size of the
67 test program. First, ``bugpoint`` figures out which combination of optimizer
107 which are causing the miscompilation, then reduce the portion of the test
108 program which is being miscompiled. The miscompilation debugger assumes that
160 * ``bugpoint`` can produce IR which contains long names. Run ``opt
175 Turn on ``-debug-only=instcombine`` and see which transformations within
[all …]
|
| /dports/lang/solidity/solidity_0.8.11/docs/ |
| H A D | yul.rst | 37 ``7 y x add mul`` because in the first form, it is much easier to see which
38 operand is used for which opcode.
57 which allows specializing Yul to the requirements of different
62 (see below) and defines only the type ``u256``, which is the native 256-bit
251 in which argument lists are evaluated. The return values,
374 Yul supports for-loops which consist of
460 which are explained in their own chapter.
1048 (the optimizer might directly re-use result of the
1064 the stack layout, which means that e.g. using ``verbatim`` to modify
1223 Please refer to the general :ref:`optimizer documentation <optimizer>`
[all …]
|
| /dports/textproc/kibana6/kibana-6.8.16-darwin-x86_64/node_modules/regexp-tree/ |
| H A D | README.md | 22 - [Using optimizer API](#using-optimizer-api)
23 - [Optimizer ESLint plugin](#optimizer-eslint-plugin)
81 -o, --optimize Applies optimizer on the passed expression
245 …d sets global options, which are preserved between calls. It is also possible to provide options p…
268 You can also provide a `*` handler which will be executed on every node.
415 ### Using optimizer API
435 From CLI the optimizer is available via `--optimize` (`-o`) option:
447 See the [optimizer README](https://github.com/DmitrySoshnikov/regexp-tree/tree/master/src/optimizer…
451 The [optimizer](https://github.com/DmitrySoshnikov/regexp-tree/tree/master/src/optimizer) module is…
863 A char preceded with `\c`, e.g. `\cx`, which stands for `CTRL+x`:
[all …]
|
| /dports/math/R-cran-lme4/lme4/R/ |
| H A D | profile.R | 26 optimizer = NULL, argument
45 if (is.null(optimizer)) optimizer <- fitted@optinfo$optimizer
94 which <- get.which(which, nvp, nptot, names(opt), verbose)
233 ores <- tryCatch(optwrap(optimizer, par=start,
361 ores <- optwrap(optimizer, par = thopt, fn = mkdevfun(rho, 0L),
418 if (is.null(which))
437 which
626 which <- get.which(which, nvp, nptot, nms)
952 which <- get.which(which, nvp, nptot, nms)
953 if (length(which) == 1)
[all …]
|
| /dports/lang/mono/mono-5.10.1.57/external/corefx/src/System.Private.Xml/src/System/Xml/Xsl/IlGen/ |
| H A D | XmlILOptimizer.p | 21 ### ignored for purposes of optimization. Nodes which are assumed to have significant side
25 ### the optimizer does not care about this side effect.
27 ### These are the rules that the optimizer follows when significant side effects are possible:
|
| /dports/devel/z88dk/z88dk/doc/ |
| H A D | copt.man | 8 copt - peephole optimizer
14 _c_o_p_t is a general-purpose peephole optimizer. It reads
17 for its optimizations, which are encoded as follows:
|
| /dports/math/R-cran-lme4/lme4/man/ |
| H A D | convergence.Rd | 14 \item adjust stopping (convergence) tolerances for the nonlinear optimizer,
33 \item the controls for the \code{nloptwrap} optimizer (the default
53 \code{rhoend}, which describes the scale of parameter uncertainty
72 …80\%93Kuhn\%E2\%80\%93Tucker_conditions}{Karush-Kuhn-Tucker conditions} (convergence criteria which
84 them by finite-difference approximations which are sometimes
131 if (all(fm1@optinfo$optimizer=="nloptwrap")) {
172 ss$ which.OK ## which fits worked
|
| /dports/finance/R-cran-tseries/tseries/man/ |
| H A D | garch.Rd | 28 \code{2*maxiter} the optimizer is allowed to compute.}
29 \item{trace}{logical. Trace optimizer output?}
45 \code{garch} uses a Quasi-Newton optimizer to find the maximum
47 max(p, q) values are assumed to be fixed. The optimizer uses a hessian
90 Two New Unconstrained Optimization Algorithms which use Function and
104 D. M. Gay, the FORTRAN optimizer
|
| /dports/www/chromium-legacy/chromium-88.0.4324.182/third_party/llvm/llvm/docs/ |
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/wasi-libcxx/llvm-project-13.0.1.src/llvm/docs/ |
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/llvm-cheri/llvm-project-37c49ff00e3eadce5d8703fdc4497f28458c64a8/llvm/docs/ |
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/llvm10/llvm-10.0.1.src/docs/ |
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/graphics/llvm-mesa/llvm-13.0.1.src/docs/ |
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/llvm12/llvm-project-12.0.1.src/llvm/docs/ |
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|
| /dports/devel/llvm11/llvm-11.0.1.src/docs/ |
| H A D | TransformMetadata.rst | 16 so, other details of how the transformations are applied (e.g., which
18 Unless the optimizer is otherwise directed, transformations are applied
19 conservatively. This conservatism generally allows the optimizer to
21 optimizer not applying transformations that would be highly profitable.
23 Frontends can give additional hints to LLVM passes on which
39 ``MDNode::replaceOperandWith`` which is dangerous to use on uniqued
54 loop distribution, etc.). They can either be a hint to the optimizer
62 ``llvm.mem.parallel_loop_access``) can be unused by the optimizer
174 ``llvm.loop.vectorize.followup_all`` which avoids that the loop
216 unrolled in which case this attribute has no effect.
[all …]
|