cross compile tensorflow c++ library for android #1
Comments
ulfjack
pushed a commit
that referenced
this issue
Jul 17, 2018
There is a fundamental bottleneck in the way targets are currently configured: The ConfiguredTargetFunction needs to load all dependent targets before it can configure its own target. Now, outgoing deps of a target can be in three states: 1. Package of dependency target not yet loaded 2. Dependency target not configured, but package already loaded 3. Dependency target configured State bazelbuild#3 is good, we don't need to do any more work before we can configure the current target. Now, most targets will have a combination of deps in states #1 and bazelbuild#2. The current behavior is to schedule the target configuration of deps in state bazelbuild#2 and the package loading in state #1 (to get to state bazelbuild#3 and bazelbuild#2, respectively). Only after *all* of these are computed will the corresponding ConfiguredTargetFunction be rescheduled upon which the remaining deps that were in state #1 before are now in state bazelbuild#2 and the corresponding target configuration can be started. This creates a substantial parallelization bottleneck by artifically sync'ing the state transitions (#1->bazelbuild#2 and bazelbuild#2->bazelbuild#3) for all deps. In turn, this makes the critical path artifically long. Instead, settle for just loading dependent packages as long as there are any. This is a fast single-step operation. Postpone the configuration of dependent targets to later as it can require long dependency chains to be configured. In theory, a better solution would be to split the target configuration in a way so that it is possible to ask for configured targets without yet knowing their package. However, that would require pulling config transitions apart (as some of them happen because of the origin target/config and some of them happen because of the destination target. Thus we'd need to change the ConfiguredTargetKey so that the configuration stands for the configuration that we request the target in, with target-caused transitions already applied. Now that itself leads to inefficiencies. E.g. we would start requestiong InputFiles in several configurations just to underneath always load them with a null configuration. RELNOTES: Faster analysis by improved parallelization. PiperOrigin-RevId: 203725209
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
i am doing cross compile tensorflow c++ library for android,and i had get the c library for android using ndk ,add like this:
cc_binary(
name = "libtensorflow.so",
srcs = [],
copts = tf_copts() + [
"-ffunction-sections",
"-fdata-sections",
],
linkopts = if_android([
"-landroid",
"-llog",
"-lm",
"-z defs",
"-s",
"-Wl,--gc-sections",
# soname is required for the so to load on api > 22
"-Wl,-soname=libtensorflow.so",
"-Wl,--version-script",
"//tensorflow/c:version_script.lds",
]),
linkshared = 1,
linkstatic = 1,
tags = [
"manual",
"notap",
],
deps = [
"//tensorflow/c:c_api",
"//tensorflow/c:version_script.lds",
"//tensorflow/core:android_tensorflow_lib",
],
)
it success.but i try c++ ,failed,and i add like this:
cc_binary(
name = "libtensorflow_cc.so",
srcs = [],
copts = tf_copts() + [
"-ffunction-sections",
"-fdata-sections",
],
linkopts = if_android([
"-landroid",
"-llog",
"-lm",
"-z defs",
"-s",
"-Wl,--gc-sections",
# soname is required for the so to load on api > 22
"-Wl,-soname=libtensorflow_cc.so",
]),
linkshared = 1,
linkstatic = 1,
tags = [
"manual",
"notap",
],
deps = [
"//tensorflow/c:c_api",
"//tensorflow/cc:cc_ops",
"//tensorflow/cc:client_session",
"//tensorflow/cc:scope",
"//tensorflow/core:android_tensorflow_lib",
],
)
it got the error:
ERROR: /home/z/e0028/workspace/tensorflow-build/target/tensorflow/tensorflow/core/BUILD:1395:1: no such target '//tensorflow/tools/git:gen/spec.json': target 'gen/spec.json' not declared in package 'tensorflow/tools/git' defined by /home/z/e0028/workspace/tensorflow-build/target/tensorflow/tensorflow/tools/git/BUILD and referenced by '//tensorflow/core:version_info_gen'
ERROR: /home/z/e0028/workspace/tensorflow-build/target/tensorflow/tensorflow/core/BUILD:1395:1: no such target '//tensorflow/tools/git:gen/head': target 'gen/head' not declared in package 'tensorflow/tools/git' defined by /home/z/e0028/workspace/tensorflow-build/target/tensorflow/tensorflow/tools/git/BUILD and referenced by '//tensorflow/core:version_info_gen'
ERROR: /home/z/e0028/workspace/tensorflow-build/target/tensorflow/tensorflow/core/BUILD:1395:1: no such target '//tensorflow/tools/git:gen/branch_ref': target 'gen/branch_ref' not declared in package 'tensorflow/tools/git' defined by /home/z/e0028/workspace/tensorflow-build/target/tensorflow/tensorflow/tools/git/BUILD and referenced by '//tensorflow/core:version_info_gen'
ERROR: Analysis of target '//tensorflow/contrib/android:libtensorflow_cc.so' failed; build aborted.
can you help me to fix it,thanks!
The text was updated successfully, but these errors were encountered: