summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Implement Make-style order-only dependenciesPeter Collingbourne2011-10-243-19/+59
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously, the implementation of order-only dependencies differed between Make and Ninja in two important ways: 1) If the order-only dependency existed but was out of date, it would never be rebuilt, whereas Make would always rebuild out of date order-only dependencies. 2) If the order-only dependency did not exist, it would cause its reverse dependencies to always build, whereas Make would only rebuild a file if a non-order-only dependency was out of date. A key distinction between Ninja and Make as seen through the above two points was that in Ninja, order-only dependencies cared about whether the target as a file exists (so perhaps a better name for the old semantics would have been "missing-only dependencies"). These differences made it impossible to introduce an order-only dependency on an always out-of-date (i.e. missing) target without also causing the depender and its reverse dependencies to rebuild unnecessarily on every build. Build systems which must perform some action (such as logging the build start time, or printing a message) at the start of every build typically implement this by adding to every target an order-only dependency which performs this action, which would have forced an entire rebuild on every invocation of Ninja under the old semantics. This commit causes Ninja to conform to the Make-style behaviour.
* Split Node::dirty_ into two flags: Node::dirty_ and Edge::outputs_ready_Peter Collingbourne2011-10-188-106/+160
| | | | | dirty_ is intended to remain static during the build (unless a restat occurs), while outputs_ready_ reflects the dynamic state of the build.
* Merge pull request #120 from pcc/generatorEvan Martin2011-10-1710-38/+94
|\ | | | | Implement generator rules
| * Make the command line interface for the clean tool consistentPeter Collingbourne2011-10-152-31/+23
| | | | | | | | | | | | Since we have started using command line flags for the clean tool, it is inconsistent to keep the "target" and "rule" prefixes. Replace them with a "-r" flag with the same semantics as "rule".
| * Implement generator rulesPeter Collingbourne2011-10-1510-19/+83
|/ | | | | | | | | | | | | Introduce a rule attribute "generator" which, if present, specifies that this rule is used to re-invoke the generator program. Files built using generator rules are treated specially in two ways: firstly, they will not be rebuilt if the command line changes; and secondly, they are not cleaned by default. A command line flag "-g" is introduced for the clean tool, which causes it to remove generator files. Fixes issue #102.
* pass extra flags into the toolEvan Martin2011-10-141-2/+5
|
* Elide output in the middle when exceeding terminal width.Nicolas Despres2011-10-121-3/+4
| | | | | | | | | | When building projects with long file paths, the rule name may disappear since the output is elided on the left side. So you no longer know whether you are compiling or linking. I think the user is interested in both the rule name and the file name. Eliding the output in the middle solves this problem.
* make CanonicalizePath report an error on empty pathEvan Martin2011-10-067-19/+57
| | | | Fixes part of issue 121, but the fix exposed a further issue.
* Merge pull request #119 from pcc/disable-assertsEvan Martin2011-09-281-1/+1
|\ | | | | Disable assertion checks unless --debug is enabled
| * Disable assertion checks unless --debug is enabledPeter Collingbourne2011-09-241-1/+1
|/
* add syntax to build output from a given fileEvan Martin2011-09-221-0/+23
| | | | e.g. "ninja src/graph.cc^" builds the object file generated from that input
* command line supports multiple targetsEvan Martin2011-09-221-2/+2
|
* make options constEvan Martin2011-09-221-6/+6
|
* stylistic cleanupsEvan Martin2011-09-221-19/+20
|
* note that target is optional in command-line helpEvan Martin2011-09-221-1/+3
|
* delete todo; I converted it into bugs on githubEvan Martin2011-09-151-22/+0
|
* delete .a files before adding to them with arEvan Martin2011-09-141-1/+1
| | | | | | | ar doesn't have a mode to just rebuild the file from scratch, so delete it manually. Fixes issue #105.
* Merge pull request #106 from pcc/depfile-vectorEvan Martin2011-09-141-4/+7
|\ | | | | Allocate space in the inputs vector for the depfile implicit deps
| * Allocate space in the inputs vector for the depfile implicit depsPeter Collingbourne2011-09-141-4/+7
| | | | | | | | This speeds up a no-op LLVM/Clang build by about 20ms.
* | Merge pull request #107 from pcc/missing-phonyEvan Martin2011-09-145-8/+40
|\ \ | |/ |/| Consider missing phony targets with no dependencies out of date
| * Consider missing phony targets with no dependencies out of datePeter Collingbourne2011-09-143-1/+19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 639c8f0 ("don't mark phony edges dirty if none of their inputs are dirty") modified the behaviour of the "phony" built-in rule. Previously, when the output file was missing, it was marked as dirty. After 639c8f0, it was always marked as clean unless one of the dependencies was dirty. The depfile mechanism uses the old behaviour of "phony" to rebuild an object file if any of the headers were missing. Restore the old "phony" behaviour only for the case where the build statement has no dependencies. This is slightly inconsistent, but I can't really see any other use case for an alias of nothing. Also, document this behaviour.
| * Modify the build tests to use the virtual file systemPeter Collingbourne2011-09-143-7/+21
|/
* generalize the pattern of hash_map<const char*, ...>, use in BuildLogEvan Martin2011-09-125-32/+42
| | | | | | Refactor the code in StatCache for use in BuildLog. Now both use hash tables where the keys are const char*. Removes another 30ms from Chrome no-op builds.
* use StringPiece for makefile depsEvan Martin2011-09-125-12/+77
| | | | | | | | | | | | Because of this, MakefileParser now returns pointers into the source makefile string rather than allocating new strings. Despite needing to take the result and stuff it into a new string anyway to canonicalize it, this takes another 50ms or so off the null Chrome build, likely due to the vector used in MakefileParser changing to a type that doesn't use any allocations. (I also experimented with making the vector reserve an initial size but didn't see any performance impact.)
* Merge pull request #104 from tfarina/ninja-testEvan Martin2011-09-124-144/+151
|\ | | | | Split ninja_test.cc into state_test.cc and disk_interface_test.cc
| * Split ninja_test.cc into state_test.cc and disk_interface_test.ccThiago Farina2011-09-114-144/+151
|/ | | | Signed-off-by: Thiago Farina <tfarina@chromium.org>
* Merge pull request #103 from tfarina/structEvan Martin2011-09-111-1/+1
|\ | | | | This is C++, there is no need to write 'struct' before the type when decl
| * This is C++, there is no need to write 'struct' before the type when ↵Thiago Farina2011-09-101-1/+1
| | | | | | | | | | | | declaring a variable. Signed-off-by: Thiago Farina <tfarina@chromium.org>
* | change main hash to char*; use pointers into filesEvan Martin2011-09-102-3/+26
| | | | | | | | | | This has a small (20ms) effect on Chrome, but it will also be used more as I remove more strings.
* | optimize IsIdentCharEvan Martin2011-09-091-5/+25
| | | | | | | | Rather than nested tests, use a table. Shaves 200ms off Chrome null startup.
* | fix "ignored return value of write" harderEvan Martin2011-09-091-2/+4
| |
* | verify -k param is numericEvan Martin2011-09-092-4/+10
| | | | | | | | Fixes issue #85.
* | minor cleanupsEvan Martin2011-09-091-2/+4
| |
* | don't mark phony edges dirty if none of their inputs are dirtyEvan Martin2011-09-092-8/+23
|/ | | | | | | Because the output file is always missing, we'd consider a phony edge dirty even when there wasn't any work to do. Most importantly, that would mean we wouldn't print "nothing to do" in the common case of everything being up to date when building an alias.
* Merge pull request #101 from tfarina/build-config-referenceEvan Martin2011-09-092-5/+5
|\ | | | | Make Cleaner::config_ a reference-to-const instead of copying the value.
| * Make Cleaner::config_ a reference-to-const instead of copying the value.Thiago Farina2011-09-092-5/+5
| | | | | | | | Signed-off-by: Thiago Farina <tfarina@chromium.org>
* | Merge pull request #95 from qhuo/depfile-for-command-with-multiple-outputsEvan Martin2011-09-093-22/+11
|\ \ | |/ |/| Add depfile support to build command with multiple outputs (Fixes: #61)
| * Move a passing test from Errors to MultipleOutputsQingning Huo2011-09-061-10/+11
| |
| * Add depfile support to build command with multiple outputs (Fixes: #61)Qingning Huo2011-09-063-14/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | parsers.cpp: allow depfile used at build command with multiple outputs. graph.cpp: allow depfile used at build command with multiple outputs. parsers_test.cpp: make the test pass. As before, the depfile itself can only mention one target, which must be the first of a build command with multiple outpus. [There is really no need to mention all the output in the depfile, because all targets should depend on exactly the same files anyway, because these targets are built by a single build command.]
* | link gtest code with -lpthreadEvan Martin2011-09-091-1/+1
| | | | | | | | I broke this in a refactoring.
* | Merge pull request #99 from tfarina/eval-env-testEvan Martin2011-09-093-75/+103
|\ \ | | | | | | Move EvalString tests to eval_env_test.cc
| * | Move EvalString tests to eval_env_test.ccThiago Farina2011-09-093-75/+103
|/ / | | | | | | Signed-off-by: Thiago Farina <tfarina@chromium.org>
* | Merge pull request #98 from tfarina/fix-win-buildEvan Martin2011-09-082-28/+29
|\ \ | | | | | | Fix windows build by moving mkdtemp() implementation from ninja_test.cc t
| * | Fix windows build by moving mkdtemp() implementation from ninja_test.cc to ↵Thiago Farina2011-09-082-28/+29
| | | | | | | | | | | | | | | | | | disk_interface_test.cc Signed-off-by: Thiago Farina <tfarina@chromium.org>
* | | Merge branch 'master' of github.com:martine/ninjaEvan Martin2011-09-083-105/+134
|\ \ \ | |/ / | | | | | | | | | Conflicts: configure.py
| * | Merge pull request #96 from tfarina/disk-interface-testEvan Martin2011-09-073-105/+134
| |\ \ | | | | | | | | Move DiskInterfaceTest into its own source file.
| | * | Move DiskInterfaceTest into its own source file.Thiago Farina2011-09-073-105/+134
| | |/ | | | | | | | | | | | | | | | TEST=ninja_test --gtest_filter=DiskInterface* Signed-off-by: Thiago Farina <tfarina@chromium.org>
* | | fix redundant manifest rebuildEvan Martin2011-09-084-20/+41
| | | | | | | | | | | | | | | | | | | | | Don't rebuild the manifest when it's already up to date. The underlying problem was that Builder::Build has a confusing API; split the API so it's more clear for callers what the return values mean.
* | | allow --with-gtest flag to use unpacked gtest dirEvan Martin2011-09-072-9/+30
|/ / | | | | | | Useful for building on Macs.
* | preserve configure.py params across re-runsEvan Martin2011-09-061-1/+4
| |
generated by cgit v0.12 at 2025-04-10 17:28:59 (GMT)