2022-02-21 19:05:27 +00:00
|
|
|
const std = @import("std");
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
const Cases = @import("src/Cases.zig");
|
2022-02-21 19:05:27 +00:00
|
|
|
|
zig build system: change target, compilation, and module APIs
Introduce the concept of "target query" and "resolved target". A target
query is what the user specifies, with some things left to default. A
resolved target has the default things discovered and populated.
In the future, std.zig.CrossTarget will be rename to std.Target.Query.
Introduces `std.Build.resolveTargetQuery` to get from one to the other.
The concept of `main_mod_path` is gone, no longer supported. You have to
put the root source file at the module root now.
* remove deprecated API
* update build.zig for the breaking API changes in this branch
* move std.Build.Step.Compile.BuildId to std.zig.BuildId
* add more options to std.Build.ExecutableOptions, std.Build.ObjectOptions,
std.Build.SharedLibraryOptions, std.Build.StaticLibraryOptions, and
std.Build.TestOptions.
* remove `std.Build.constructCMacro`. There is no use for this API.
* deprecate `std.Build.Step.Compile.defineCMacro`. Instead,
`std.Build.Module.addCMacro` is provided.
- remove `std.Build.Step.Compile.defineCMacroRaw`.
* deprecate `std.Build.Step.Compile.linkFrameworkNeeded`
- use `std.Build.Module.linkFramework`
* deprecate `std.Build.Step.Compile.linkFrameworkWeak`
- use `std.Build.Module.linkFramework`
* move more logic into `std.Build.Module`
* allow `target` and `optimize` to be `null` when creating a Module.
Along with other fields, those unspecified options will be inherited
from parent `Module` when inserted into an import table.
* the `target` field of `addExecutable` is now required. pass `b.host`
to get the host target.
2023-12-03 04:51:34 +00:00
|
|
|
pub fn addCases(ctx: *Cases, b: *std.Build) !void {
|
|
|
|
|
const target = b.resolveTargetQuery(.{
|
|
|
|
|
.cpu_arch = .nvptx64,
|
|
|
|
|
.os_tag = .cuda,
|
|
|
|
|
});
|
|
|
|
|
|
2022-02-21 19:05:27 +00:00
|
|
|
{
|
zig build system: change target, compilation, and module APIs
Introduce the concept of "target query" and "resolved target". A target
query is what the user specifies, with some things left to default. A
resolved target has the default things discovered and populated.
In the future, std.zig.CrossTarget will be rename to std.Target.Query.
Introduces `std.Build.resolveTargetQuery` to get from one to the other.
The concept of `main_mod_path` is gone, no longer supported. You have to
put the root source file at the module root now.
* remove deprecated API
* update build.zig for the breaking API changes in this branch
* move std.Build.Step.Compile.BuildId to std.zig.BuildId
* add more options to std.Build.ExecutableOptions, std.Build.ObjectOptions,
std.Build.SharedLibraryOptions, std.Build.StaticLibraryOptions, and
std.Build.TestOptions.
* remove `std.Build.constructCMacro`. There is no use for this API.
* deprecate `std.Build.Step.Compile.defineCMacro`. Instead,
`std.Build.Module.addCMacro` is provided.
- remove `std.Build.Step.Compile.defineCMacroRaw`.
* deprecate `std.Build.Step.Compile.linkFrameworkNeeded`
- use `std.Build.Module.linkFramework`
* deprecate `std.Build.Step.Compile.linkFrameworkWeak`
- use `std.Build.Module.linkFramework`
* move more logic into `std.Build.Module`
* allow `target` and `optimize` to be `null` when creating a Module.
Along with other fields, those unspecified options will be inherited
from parent `Module` when inserted into an import table.
* the `target` field of `addExecutable` is now required. pass `b.host`
to get the host target.
2023-12-03 04:51:34 +00:00
|
|
|
var case = addPtx(ctx, target, "simple addition and subtraction");
|
2022-02-21 19:05:27 +00:00
|
|
|
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
case.addCompile(
|
2022-02-21 19:05:27 +00:00
|
|
|
\\fn add(a: i32, b: i32) i32 {
|
|
|
|
|
\\ return a + b;
|
|
|
|
|
\\}
|
|
|
|
|
\\
|
2023-04-08 13:17:34 +01:00
|
|
|
\\pub export fn add_and_substract(a: i32, out: *i32) callconv(.Kernel) void {
|
2022-02-21 19:05:27 +00:00
|
|
|
\\ const x = add(a, 7);
|
|
|
|
|
\\ var y = add(2, 0);
|
|
|
|
|
\\ y -= x;
|
|
|
|
|
\\ out.* = y;
|
|
|
|
|
\\}
|
|
|
|
|
);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
{
|
zig build system: change target, compilation, and module APIs
Introduce the concept of "target query" and "resolved target". A target
query is what the user specifies, with some things left to default. A
resolved target has the default things discovered and populated.
In the future, std.zig.CrossTarget will be rename to std.Target.Query.
Introduces `std.Build.resolveTargetQuery` to get from one to the other.
The concept of `main_mod_path` is gone, no longer supported. You have to
put the root source file at the module root now.
* remove deprecated API
* update build.zig for the breaking API changes in this branch
* move std.Build.Step.Compile.BuildId to std.zig.BuildId
* add more options to std.Build.ExecutableOptions, std.Build.ObjectOptions,
std.Build.SharedLibraryOptions, std.Build.StaticLibraryOptions, and
std.Build.TestOptions.
* remove `std.Build.constructCMacro`. There is no use for this API.
* deprecate `std.Build.Step.Compile.defineCMacro`. Instead,
`std.Build.Module.addCMacro` is provided.
- remove `std.Build.Step.Compile.defineCMacroRaw`.
* deprecate `std.Build.Step.Compile.linkFrameworkNeeded`
- use `std.Build.Module.linkFramework`
* deprecate `std.Build.Step.Compile.linkFrameworkWeak`
- use `std.Build.Module.linkFramework`
* move more logic into `std.Build.Module`
* allow `target` and `optimize` to be `null` when creating a Module.
Along with other fields, those unspecified options will be inherited
from parent `Module` when inserted into an import table.
* the `target` field of `addExecutable` is now required. pass `b.host`
to get the host target.
2023-12-03 04:51:34 +00:00
|
|
|
var case = addPtx(ctx, target, "read special registers");
|
2022-02-21 19:05:27 +00:00
|
|
|
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
case.addCompile(
|
2022-10-04 06:31:36 +01:00
|
|
|
\\fn threadIdX() u32 {
|
|
|
|
|
\\ return asm ("mov.u32 \t%[r], %tid.x;"
|
2022-10-06 21:06:20 +01:00
|
|
|
\\ : [r] "=r" (-> u32),
|
2022-10-04 06:31:36 +01:00
|
|
|
\\ );
|
2022-02-21 19:05:27 +00:00
|
|
|
\\}
|
|
|
|
|
\\
|
2023-04-08 13:17:34 +01:00
|
|
|
\\pub export fn special_reg(a: []const i32, out: []i32) callconv(.Kernel) void {
|
2022-03-01 22:26:43 +00:00
|
|
|
\\ const i = threadIdX();
|
2022-02-21 19:05:27 +00:00
|
|
|
\\ out[i] = a[i] + 7;
|
|
|
|
|
\\}
|
|
|
|
|
);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
{
|
zig build system: change target, compilation, and module APIs
Introduce the concept of "target query" and "resolved target". A target
query is what the user specifies, with some things left to default. A
resolved target has the default things discovered and populated.
In the future, std.zig.CrossTarget will be rename to std.Target.Query.
Introduces `std.Build.resolveTargetQuery` to get from one to the other.
The concept of `main_mod_path` is gone, no longer supported. You have to
put the root source file at the module root now.
* remove deprecated API
* update build.zig for the breaking API changes in this branch
* move std.Build.Step.Compile.BuildId to std.zig.BuildId
* add more options to std.Build.ExecutableOptions, std.Build.ObjectOptions,
std.Build.SharedLibraryOptions, std.Build.StaticLibraryOptions, and
std.Build.TestOptions.
* remove `std.Build.constructCMacro`. There is no use for this API.
* deprecate `std.Build.Step.Compile.defineCMacro`. Instead,
`std.Build.Module.addCMacro` is provided.
- remove `std.Build.Step.Compile.defineCMacroRaw`.
* deprecate `std.Build.Step.Compile.linkFrameworkNeeded`
- use `std.Build.Module.linkFramework`
* deprecate `std.Build.Step.Compile.linkFrameworkWeak`
- use `std.Build.Module.linkFramework`
* move more logic into `std.Build.Module`
* allow `target` and `optimize` to be `null` when creating a Module.
Along with other fields, those unspecified options will be inherited
from parent `Module` when inserted into an import table.
* the `target` field of `addExecutable` is now required. pass `b.host`
to get the host target.
2023-12-03 04:51:34 +00:00
|
|
|
var case = addPtx(ctx, target, "address spaces");
|
2022-02-21 19:05:27 +00:00
|
|
|
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
case.addCompile(
|
2022-03-01 22:26:43 +00:00
|
|
|
\\var x: i32 addrspace(.global) = 0;
|
2022-02-21 19:05:27 +00:00
|
|
|
\\
|
2023-04-08 13:17:34 +01:00
|
|
|
\\pub export fn increment(out: *i32) callconv(.Kernel) void {
|
2022-02-21 19:05:27 +00:00
|
|
|
\\ x += 1;
|
|
|
|
|
\\ out.* = x;
|
|
|
|
|
\\}
|
|
|
|
|
);
|
|
|
|
|
}
|
2022-10-04 06:31:36 +01:00
|
|
|
|
|
|
|
|
{
|
zig build system: change target, compilation, and module APIs
Introduce the concept of "target query" and "resolved target". A target
query is what the user specifies, with some things left to default. A
resolved target has the default things discovered and populated.
In the future, std.zig.CrossTarget will be rename to std.Target.Query.
Introduces `std.Build.resolveTargetQuery` to get from one to the other.
The concept of `main_mod_path` is gone, no longer supported. You have to
put the root source file at the module root now.
* remove deprecated API
* update build.zig for the breaking API changes in this branch
* move std.Build.Step.Compile.BuildId to std.zig.BuildId
* add more options to std.Build.ExecutableOptions, std.Build.ObjectOptions,
std.Build.SharedLibraryOptions, std.Build.StaticLibraryOptions, and
std.Build.TestOptions.
* remove `std.Build.constructCMacro`. There is no use for this API.
* deprecate `std.Build.Step.Compile.defineCMacro`. Instead,
`std.Build.Module.addCMacro` is provided.
- remove `std.Build.Step.Compile.defineCMacroRaw`.
* deprecate `std.Build.Step.Compile.linkFrameworkNeeded`
- use `std.Build.Module.linkFramework`
* deprecate `std.Build.Step.Compile.linkFrameworkWeak`
- use `std.Build.Module.linkFramework`
* move more logic into `std.Build.Module`
* allow `target` and `optimize` to be `null` when creating a Module.
Along with other fields, those unspecified options will be inherited
from parent `Module` when inserted into an import table.
* the `target` field of `addExecutable` is now required. pass `b.host`
to get the host target.
2023-12-03 04:51:34 +00:00
|
|
|
var case = addPtx(ctx, target, "reduce in shared mem");
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
case.addCompile(
|
2022-10-04 06:31:36 +01:00
|
|
|
\\fn threadIdX() u32 {
|
|
|
|
|
\\ return asm ("mov.u32 \t%[r], %tid.x;"
|
2022-10-06 21:06:20 +01:00
|
|
|
\\ : [r] "=r" (-> u32),
|
2022-10-04 06:31:36 +01:00
|
|
|
\\ );
|
|
|
|
|
\\}
|
|
|
|
|
\\
|
|
|
|
|
\\ var _sdata: [1024]f32 addrspace(.shared) = undefined;
|
2023-04-08 13:17:34 +01:00
|
|
|
\\ pub export fn reduceSum(d_x: []const f32, out: *f32) callconv(.Kernel) void {
|
2023-06-22 18:46:56 +01:00
|
|
|
\\ var sdata: *addrspace(.generic) [1024]f32 = @addrSpaceCast(&_sdata);
|
2022-10-04 06:31:36 +01:00
|
|
|
\\ const tid: u32 = threadIdX();
|
|
|
|
|
\\ var sum = d_x[tid];
|
|
|
|
|
\\ sdata[tid] = sum;
|
|
|
|
|
\\ asm volatile ("bar.sync \t0;");
|
|
|
|
|
\\ var s: u32 = 512;
|
|
|
|
|
\\ while (s > 0) : (s = s >> 1) {
|
|
|
|
|
\\ if (tid < s) {
|
|
|
|
|
\\ sum += sdata[tid + s];
|
|
|
|
|
\\ sdata[tid] = sum;
|
|
|
|
|
\\ }
|
|
|
|
|
\\ asm volatile ("bar.sync \t0;");
|
|
|
|
|
\\ }
|
|
|
|
|
\\
|
|
|
|
|
\\ if (tid == 0) {
|
|
|
|
|
\\ out.* = sum;
|
|
|
|
|
\\ }
|
|
|
|
|
\\ }
|
|
|
|
|
);
|
|
|
|
|
}
|
2022-02-21 19:05:27 +00:00
|
|
|
}
|
2022-03-01 22:26:43 +00:00
|
|
|
|
zig build system: change target, compilation, and module APIs
Introduce the concept of "target query" and "resolved target". A target
query is what the user specifies, with some things left to default. A
resolved target has the default things discovered and populated.
In the future, std.zig.CrossTarget will be rename to std.Target.Query.
Introduces `std.Build.resolveTargetQuery` to get from one to the other.
The concept of `main_mod_path` is gone, no longer supported. You have to
put the root source file at the module root now.
* remove deprecated API
* update build.zig for the breaking API changes in this branch
* move std.Build.Step.Compile.BuildId to std.zig.BuildId
* add more options to std.Build.ExecutableOptions, std.Build.ObjectOptions,
std.Build.SharedLibraryOptions, std.Build.StaticLibraryOptions, and
std.Build.TestOptions.
* remove `std.Build.constructCMacro`. There is no use for this API.
* deprecate `std.Build.Step.Compile.defineCMacro`. Instead,
`std.Build.Module.addCMacro` is provided.
- remove `std.Build.Step.Compile.defineCMacroRaw`.
* deprecate `std.Build.Step.Compile.linkFrameworkNeeded`
- use `std.Build.Module.linkFramework`
* deprecate `std.Build.Step.Compile.linkFrameworkWeak`
- use `std.Build.Module.linkFramework`
* move more logic into `std.Build.Module`
* allow `target` and `optimize` to be `null` when creating a Module.
Along with other fields, those unspecified options will be inherited
from parent `Module` when inserted into an import table.
* the `target` field of `addExecutable` is now required. pass `b.host`
to get the host target.
2023-12-03 04:51:34 +00:00
|
|
|
fn addPtx(ctx: *Cases, target: std.Build.ResolvedTarget, name: []const u8) *Cases.Case {
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
ctx.cases.append(.{
|
2022-03-01 22:26:43 +00:00
|
|
|
.name = name,
|
zig build system: change target, compilation, and module APIs
Introduce the concept of "target query" and "resolved target". A target
query is what the user specifies, with some things left to default. A
resolved target has the default things discovered and populated.
In the future, std.zig.CrossTarget will be rename to std.Target.Query.
Introduces `std.Build.resolveTargetQuery` to get from one to the other.
The concept of `main_mod_path` is gone, no longer supported. You have to
put the root source file at the module root now.
* remove deprecated API
* update build.zig for the breaking API changes in this branch
* move std.Build.Step.Compile.BuildId to std.zig.BuildId
* add more options to std.Build.ExecutableOptions, std.Build.ObjectOptions,
std.Build.SharedLibraryOptions, std.Build.StaticLibraryOptions, and
std.Build.TestOptions.
* remove `std.Build.constructCMacro`. There is no use for this API.
* deprecate `std.Build.Step.Compile.defineCMacro`. Instead,
`std.Build.Module.addCMacro` is provided.
- remove `std.Build.Step.Compile.defineCMacroRaw`.
* deprecate `std.Build.Step.Compile.linkFrameworkNeeded`
- use `std.Build.Module.linkFramework`
* deprecate `std.Build.Step.Compile.linkFrameworkWeak`
- use `std.Build.Module.linkFramework`
* move more logic into `std.Build.Module`
* allow `target` and `optimize` to be `null` when creating a Module.
Along with other fields, those unspecified options will be inherited
from parent `Module` when inserted into an import table.
* the `target` field of `addExecutable` is now required. pass `b.host`
to get the host target.
2023-12-03 04:51:34 +00:00
|
|
|
.target = target,
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
.updates = std.ArrayList(Cases.Update).init(ctx.cases.allocator),
|
2022-03-01 22:26:43 +00:00
|
|
|
.output_mode = .Obj,
|
re-enable test-cases and get them all passing
Instead of using `zig test` to build a special version of the compiler
that runs all the test-cases, the zig build system is now used as much
as possible - all with the basic steps found in the standard library.
For incremental compilation tests (the ones that look like foo.0.zig,
foo.1.zig, foo.2.zig, etc.), a special version of the compiler is
compiled into a utility executable called "check-case" which checks
exactly one sequence of incremental updates in an independent
subprocess. Previously, all incremental and non-incremental test cases
were done in the same test runner process.
The compile error checking code is now simpler, but also a bit
rudimentary, and so it additionally makes sure that the actual compile
errors do not include *extra* messages, and it makes sure that the
actual compile errors output in the same order as expected. It is also
based on the "ends-with" property of each line rather than the previous
logic, which frankly I didn't want to touch with a ten-meter pole. The
compile error test cases have been updated to pass in light of these
differences.
Previously, 'error' mode with 0 compile errors was used to shoehorn in a
different kind of test-case - one that only checks if a piece of code
compiles without errors. Now there is a 'compile' mode of test-cases,
and 'error' must be only used when there are greater than 0 errors.
link test cases are updated to omit the target object format argument
when calling checkObject since that is no longer needed.
The test/stage2 directory is removed; the 2 files within are moved to be
directly in the test/ directory.
2023-03-10 01:22:51 +00:00
|
|
|
.deps = std.ArrayList(Cases.DepModule).init(ctx.cases.allocator),
|
2022-03-01 22:26:43 +00:00
|
|
|
.link_libc = false,
|
|
|
|
|
.backend = .llvm,
|
2022-10-06 21:06:20 +01:00
|
|
|
// Bug in Debug mode
|
|
|
|
|
.optimize_mode = .ReleaseSafe,
|
2022-03-01 22:26:43 +00:00
|
|
|
}) catch @panic("out of memory");
|
|
|
|
|
return &ctx.cases.items[ctx.cases.items.len - 1];
|
|
|
|
|
}
|
