Merge pull request #109 from klueska/move-dev-to-pkg

Move development directly into pkg/nvml instead of gen/nvml

.gitignore

@@ -1,3 +1,4 @@
 *.swp
 *.swo
-pkg/nvml/
+.cache
+coverage.out*

Makefile

@@ -139,12 +139,11 @@ bindings: .create-bindings .strip-autogen-comment .strip-nvml-h-linenumber
 .create-bindings: $(PKG_BINDINGS_DIR)/nvml.h $(SOURCES) | $(PKG_BINDINGS_DIR)
 	cp $(GEN_BINDINGS_DIR)/nvml.yml $(PKG_BINDINGS_DIR)
 	c-for-go -out $(PKG_DIR) $(PKG_BINDINGS_DIR)/nvml.yml
-	cp $(GEN_BINDINGS_DIR)/*.go $(PKG_BINDINGS_DIR)
 	cd $(PKG_BINDINGS_DIR); \
 		go tool cgo -godefs types.go > types_gen.go; \
 		go fmt types_gen.go; \
 	cd -> /dev/null
-	rm -rf $(PKG_BINDINGS_DIR)/nvml.yml $(PKG_BINDINGS_DIR)/types.go $(PKG_BINDINGS_DIR)/_obj
+	rm -rf $(PKG_BINDINGS_DIR)/nvml.yml $(PKG_BINDINGS_DIR)/cgo_helpers.go $(PKG_BINDINGS_DIR)/types.go $(PKG_BINDINGS_DIR)/_obj
 
 .strip-autogen-comment: SED_SEARCH_STRING := // WARNING: This file has automatically been generated on
 .strip-autogen-comment: SED_REPLACE_STRING := // WARNING: THIS FILE WAS AUTOMATICALLY GENERATED.
@@ -160,12 +159,16 @@ bindings: .create-bindings .strip-autogen-comment .strip-nvml-h-linenumber
 
 test-bindings: bindings
 clean-bindings:
-	rm -rf $(PKG_BINDINGS_DIR)
-	git checkout $(PKG_BINDINGS_DIR)
-	rm -rf $(PKG_BINDINGS_DIR)/nvml.h
+	rm -f $(PKG_BINDINGS_DIR)/cgo_helpers.go
+	rm -f $(PKG_BINDINGS_DIR)/cgo_helpers.h
+	rm -f $(PKG_BINDINGS_DIR)/const.go
+	rm -f $(PKG_BINDINGS_DIR)/doc.go
+	rm -f $(PKG_BINDINGS_DIR)/nvml.go
+	rm -f $(PKG_BINDINGS_DIR)/nvml.h
+	rm -f $(PKG_BINDINGS_DIR)/types_gen.go
 
 # Update nvml.h from the Anaconda package repository
-update-nvml-h: JQ ?= $(DOCKER) run -i --rm -v "$(PWD):$(PWD)" -w "$(PWD)" stedolan/jq:latest
+update-nvml-h: JQ ?= $(DOCKER) run -i --rm -v "$(PWD):$(PWD)" -w "$(PWD)" backplane/jq:latest
 update-nvml-h: NVML_DEV_PACKAGES_INFO := $(shell \
 		wget -qO - https://api.anaconda.org/package/nvidia/cuda-nvml-dev/files | \
 			$(JQ) '.[] | select(.attrs.subdir=="linux-64") | .version + "@" + .upload_time[:19] + "@" + .full_name' | \

README.md

@@ -321,7 +321,7 @@ symbols include the following (as defined in `nvml.h`):
 The actual versions that these API calls are assigned to will depend on the
 version of the NVIDIA driver (and hence the version of `libnvidia-ml.so` that
 you have linked in). These updates happen in the `updateVersionedSymbols()`
-function of `gen/nvml/lib.go` as seen below.
+function of `pkg/nvml/lib.go` as seen below.
 
 ```go
 // Default all versioned APIs to v1 (to infer the types)
@@ -366,8 +366,8 @@ The files below define a set of "glue" code between the auto-generated bindings
 from `c-for-go` and the manual wrappers providing a more user-friendly API to
 the end user.
 
-- `gen/nvml/cgo_helpers.go`
-- `gen/nvml/return.go`
+- `pkg/nvml/cgo_helpers_atatic.go`
+- `pkg/nvml/return.go`
 
 The `cgo_helpers.go` file defines functions that help in dealing with the types
 coming out of the C API and turning them into more usable Go types. It is
@@ -391,12 +391,12 @@ bindings from `c-for-go`. Only these manual wrappers are expected as part of
 the API for the package -- the auto-generated bindings are only available for
 internal use.
 
-- `gen/nvml/init.go`
-- `gen/nvml/system.go`
-- `gen/nvml/event_set.go`
-- `gen/nvml/vgpu.go`
-- `gen/nvml/unit.go`
-- `gen/nvml/device.go`
+- `pkg/nvml/init.go`
+- `pkg/nvml/system.go`
+- `pkg/nvml/event_set.go`
+- `pkg/nvml/vgpu.go`
+- `pkg/nvml/unit.go`
+- `pkg/nvml/device.go`
 
 These wrappers add boiler-plate code around the auto-generated bindings so that
 the end-user doesn't have to do this themselves every time a call is made.
@@ -475,7 +475,7 @@ should be straightforward so long as we keep good pace with each new release.
 
 At present, all test code is under the following file:
 
-- `gen/nvml/nvml_test.go`
+- `pkg/nvml/nvml_test.go`
 
 The test coverage is fairly sparse and could be greatly improved.