diff --git a/go.mod b/go.mod index ffafa681..3650b7a3 100644 --- a/go.mod +++ b/go.mod @@ -4,7 +4,7 @@ go 1.15 require ( cloud.google.com/go v0.74.0 // indirect - github.com/TwinProduction/gocache v1.2.0 + github.com/TwinProduction/gocache v1.2.1 github.com/go-ping/ping v0.0.0-20201115131931-3300c582a663 github.com/google/gofuzz v1.2.0 // indirect github.com/gorilla/mux v1.8.0 diff --git a/go.sum b/go.sum index 81cd012e..afb90f86 100644 --- a/go.sum +++ b/go.sum @@ -50,8 +50,8 @@ github.com/PuerkitoBio/purell v1.0.0/go.mod h1:c11w/QuzBsJSee3cPx9rAFu61PvFxuPbt github.com/PuerkitoBio/urlesc v0.0.0-20160726150825-5bd2802263f2/go.mod h1:uGdkoq3SwY9Y+13GIhn11/XLaGBb4BfwItxLd5jeuXE= github.com/Shopify/sarama v1.19.0/go.mod h1:FVkBWblsNy7DGZRfXLU0O9RCGt5g3g3yEuWXgklEdEo= github.com/Shopify/toxiproxy v2.1.4+incompatible/go.mod h1:OXgGpZ6Cli1/URJOF1DMxUHB2q5Ap20/P/eIdh4G0pI= -github.com/TwinProduction/gocache v1.2.0 h1:iZBUuri5VydxYhNkWEOZm/JzX/X2b3OZzfLrPaRWKjk= -github.com/TwinProduction/gocache v1.2.0/go.mod h1:+qH57V/K4oAcX9C7CvgJTwUX4lzfIUXQC/6XaRSOS1Y= +github.com/TwinProduction/gocache v1.2.1 h1:NAdMwO9SQEZFmX69YWx6fzhwb6fHakkLri0451c+V1w= +github.com/TwinProduction/gocache v1.2.1/go.mod h1:6zkBoLjrFLkIISwkZTgLy67qliCGSon1xpORM4Ri5HM= github.com/VividCortex/gohistogram v1.0.0/go.mod h1:Pf5mBqqDxYaXu3hDrrU+w6nw50o/4+TcAqDqk/vUH7g= github.com/afex/hystrix-go v0.0.0-20180502004556-fa1af6a1f4f5/go.mod h1:SkGFH1ia65gfNATL8TAiHDNxPzPdmEL5uirI2Uyuz6c= github.com/alecthomas/template v0.0.0-20160405071501-a0175ee3bccc/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc= @@ -73,8 +73,6 @@ github.com/beorn7/perks v1.0.0/go.mod h1:KWe93zE9D1o94FZ5RNwFwVgaQK1VOXiVxmqh+Ce github.com/beorn7/perks v1.0.1 h1:VlbKKnNfV8bJzeqoa4cOKqO6bYr3WgKZxO8Z16+hsOM= github.com/beorn7/perks v1.0.1/go.mod h1:G2ZrVWU2WbWT9wwq4/hrbKbnv/1ERSJQ0ibhJ6rlkpw= github.com/bgentry/speakeasy v0.1.0/go.mod h1:+zsyZBPWlz7T6j88CTgSN5bM796AkVf0kBD4zp0CCIs= -github.com/boltdb/bolt v1.3.1 h1:JQmyP4ZBrce+ZQu0dY660FMfatumYDLun9hBCUVIkF4= -github.com/boltdb/bolt v1.3.1/go.mod h1:clJnj/oiGkjum5o1McbSZDSLxVThjynRyGBgiAx27Ps= github.com/casbin/casbin/v2 v2.1.2/go.mod h1:YcPU1XXisHhLzuxH9coDNf2FbKpjGlbCg3n9yuLkIJQ= github.com/cenkalti/backoff v2.2.1+incompatible/go.mod h1:90ReRw6GdpyfrHakVjL/QHaoyV4aDUVVkXQJJJ3NXXM= github.com/census-instrumentation/opencensus-proto v0.2.1/go.mod h1:f6KPmirojxKA12rnyqOA5BBL4O983OfeGPqjHWSTneU= @@ -432,6 +430,8 @@ github.com/yuin/goldmark v1.1.27/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9de github.com/yuin/goldmark v1.1.32/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74= github.com/yuin/goldmark v1.2.1/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74= go.etcd.io/bbolt v1.3.3/go.mod h1:IbVyRI1SCnLcuJnV2u8VeU0CEYM7e686BmAb1XKL+uU= +go.etcd.io/bbolt v1.3.5 h1:XAzx9gjCb0Rxj7EoqcClPD1d5ZBxZJk0jbuoPHenBt0= +go.etcd.io/bbolt v1.3.5/go.mod h1:G5EMThwa9y8QZGBClrRx5EY+Yw9kAhnjy3bSjsnlVTQ= go.etcd.io/etcd v0.0.0-20191023171146-3cf2f69b5738/go.mod h1:dnLIgRNXwCJa5e+c6mIZCrds/GIG4ncV9HhK5PX7jPg= go.opencensus.io v0.20.1/go.mod h1:6WKK9ahsWS3RSO+PY9ZHZUfv2irvY6gN279GOPZjmmk= go.opencensus.io v0.20.2/go.mod h1:6WKK9ahsWS3RSO+PY9ZHZUfv2irvY6gN279GOPZjmmk= diff --git a/vendor/github.com/TwinProduction/gocache/go.mod b/vendor/github.com/TwinProduction/gocache/go.mod index bb4983b2..35cf7fd5 100644 --- a/vendor/github.com/TwinProduction/gocache/go.mod +++ b/vendor/github.com/TwinProduction/gocache/go.mod @@ -3,9 +3,9 @@ module github.com/TwinProduction/gocache go 1.15 require ( - github.com/boltdb/bolt v1.3.1 github.com/go-redis/redis v6.15.9+incompatible github.com/onsi/ginkgo v1.14.1 // indirect github.com/onsi/gomega v1.10.2 // indirect github.com/tidwall/redcon v1.3.2 + go.etcd.io/bbolt v1.3.5 ) diff --git a/vendor/github.com/TwinProduction/gocache/go.sum b/vendor/github.com/TwinProduction/gocache/go.sum index aeba13e0..5c445899 100644 --- a/vendor/github.com/TwinProduction/gocache/go.sum +++ b/vendor/github.com/TwinProduction/gocache/go.sum @@ -1,5 +1,3 @@ -github.com/boltdb/bolt v1.3.1 h1:JQmyP4ZBrce+ZQu0dY660FMfatumYDLun9hBCUVIkF4= -github.com/boltdb/bolt v1.3.1/go.mod h1:clJnj/oiGkjum5o1McbSZDSLxVThjynRyGBgiAx27Ps= github.com/fsnotify/fsnotify v1.4.7/go.mod h1:jwhsz4b93w/PPRr/qN1Yymfu8t87LnFCMoQvtojpjFo= github.com/fsnotify/fsnotify v1.4.9 h1:hsms1Qyu0jgnwNXIxa+/V/PDsU6CfLf6CNO8H7IWoS4= github.com/fsnotify/fsnotify v1.4.9/go.mod h1:znqG4EE+3YCdAaPaxE2ZRY/06pZUdp0tY4IgpuI1SZQ= @@ -30,6 +28,8 @@ github.com/onsi/gomega v1.10.2 h1:aY/nuoWlKJud2J6U0E3NWsjlg+0GtwXxgEqthRdzlcs= github.com/onsi/gomega v1.10.2/go.mod h1:iN09h71vgCQne3DLsj+A5owkum+a2tYe+TOCB1ybHNo= github.com/tidwall/redcon v1.3.2 h1:8INx/Nm3VSUbDUT16TH1rMgYQsbXNqy9xcX70edHXbo= github.com/tidwall/redcon v1.3.2/go.mod h1:bdYBm4rlcWpst2XMwKVzWDF9CoUxEbUmM7CQrKeOZas= +go.etcd.io/bbolt v1.3.5 h1:XAzx9gjCb0Rxj7EoqcClPD1d5ZBxZJk0jbuoPHenBt0= +go.etcd.io/bbolt v1.3.5/go.mod h1:G5EMThwa9y8QZGBClrRx5EY+Yw9kAhnjy3bSjsnlVTQ= golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w= golang.org/x/net v0.0.0-20180906233101-161cd47e91fd/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4= golang.org/x/net v0.0.0-20200520004742-59133d7f0dd7 h1:AeiKBIuRw3UomYXSbLy0Mc2dDLfdtbT/IVn4keq83P0= @@ -40,6 +40,7 @@ golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5h golang.org/x/sys v0.0.0-20190904154756-749cb33beabd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= golang.org/x/sys v0.0.0-20191005200804-aed5e4c7ecf9/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= golang.org/x/sys v0.0.0-20191120155948-bd437916bb0e/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= +golang.org/x/sys v0.0.0-20200202164722-d101bd2416d5/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= golang.org/x/sys v0.0.0-20200323222414-85ca7c5b95cd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= golang.org/x/sys v0.0.0-20200519105757-fe76b779f299 h1:DYfZAGf2WMFjMxbgTjaC+2HC7NkNAQs+6Q8b9WEB/F4= golang.org/x/sys v0.0.0-20200519105757-fe76b779f299/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= diff --git a/vendor/github.com/TwinProduction/gocache/persistence.go b/vendor/github.com/TwinProduction/gocache/persistence.go index 8115ebbc..55415992 100644 --- a/vendor/github.com/TwinProduction/gocache/persistence.go +++ b/vendor/github.com/TwinProduction/gocache/persistence.go @@ -8,7 +8,7 @@ import ( "sort" "time" - "github.com/boltdb/bolt" + bolt "go.etcd.io/bbolt" ) // SaveToFile stores the content of the cache to a file so that it can be read using diff --git a/vendor/github.com/boltdb/bolt/Makefile b/vendor/github.com/boltdb/bolt/Makefile deleted file mode 100644 index e035e63a..00000000 --- a/vendor/github.com/boltdb/bolt/Makefile +++ /dev/null @@ -1,18 +0,0 @@ -BRANCH=`git rev-parse --abbrev-ref HEAD` -COMMIT=`git rev-parse --short HEAD` -GOLDFLAGS="-X main.branch $(BRANCH) -X main.commit $(COMMIT)" - -default: build - -race: - @go test -v -race -test.run="TestSimulate_(100op|1000op)" - -# go get github.com/kisielk/errcheck -errcheck: - @errcheck -ignorepkg=bytes -ignore=os:Remove github.com/boltdb/bolt - -test: - @go test -v -cover . - @go test -v ./cmd/bolt - -.PHONY: fmt test diff --git a/vendor/github.com/boltdb/bolt/appveyor.yml b/vendor/github.com/boltdb/bolt/appveyor.yml deleted file mode 100644 index 6e26e941..00000000 --- a/vendor/github.com/boltdb/bolt/appveyor.yml +++ /dev/null @@ -1,18 +0,0 @@ -version: "{build}" - -os: Windows Server 2012 R2 - -clone_folder: c:\gopath\src\github.com\boltdb\bolt - -environment: - GOPATH: c:\gopath - -install: - - echo %PATH% - - echo %GOPATH% - - go version - - go env - - go get -v -t ./... - -build_script: - - go test -v ./... diff --git a/vendor/github.com/boltdb/bolt/bolt_arm.go b/vendor/github.com/boltdb/bolt/bolt_arm.go deleted file mode 100644 index 7e5cb4b9..00000000 --- a/vendor/github.com/boltdb/bolt/bolt_arm.go +++ /dev/null @@ -1,28 +0,0 @@ -package bolt - -import "unsafe" - -// maxMapSize represents the largest mmap size supported by Bolt. -const maxMapSize = 0x7FFFFFFF // 2GB - -// maxAllocSize is the size used when creating array pointers. -const maxAllocSize = 0xFFFFFFF - -// Are unaligned load/stores broken on this arch? -var brokenUnaligned bool - -func init() { - // Simple check to see whether this arch handles unaligned load/stores - // correctly. - - // ARM9 and older devices require load/stores to be from/to aligned - // addresses. If not, the lower 2 bits are cleared and that address is - // read in a jumbled up order. - - // See http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.faqs/ka15414.html - - raw := [6]byte{0xfe, 0xef, 0x11, 0x22, 0x22, 0x11} - val := *(*uint32)(unsafe.Pointer(uintptr(unsafe.Pointer(&raw)) + 2)) - - brokenUnaligned = val != 0x11222211 -} diff --git a/vendor/github.com/boltdb/bolt/freelist.go b/vendor/github.com/boltdb/bolt/freelist.go deleted file mode 100644 index aba48f58..00000000 --- a/vendor/github.com/boltdb/bolt/freelist.go +++ /dev/null @@ -1,252 +0,0 @@ -package bolt - -import ( - "fmt" - "sort" - "unsafe" -) - -// freelist represents a list of all pages that are available for allocation. -// It also tracks pages that have been freed but are still in use by open transactions. -type freelist struct { - ids []pgid // all free and available free page ids. - pending map[txid][]pgid // mapping of soon-to-be free page ids by tx. - cache map[pgid]bool // fast lookup of all free and pending page ids. -} - -// newFreelist returns an empty, initialized freelist. -func newFreelist() *freelist { - return &freelist{ - pending: make(map[txid][]pgid), - cache: make(map[pgid]bool), - } -} - -// size returns the size of the page after serialization. -func (f *freelist) size() int { - n := f.count() - if n >= 0xFFFF { - // The first element will be used to store the count. See freelist.write. - n++ - } - return pageHeaderSize + (int(unsafe.Sizeof(pgid(0))) * n) -} - -// count returns count of pages on the freelist -func (f *freelist) count() int { - return f.free_count() + f.pending_count() -} - -// free_count returns count of free pages -func (f *freelist) free_count() int { - return len(f.ids) -} - -// pending_count returns count of pending pages -func (f *freelist) pending_count() int { - var count int - for _, list := range f.pending { - count += len(list) - } - return count -} - -// copyall copies into dst a list of all free ids and all pending ids in one sorted list. -// f.count returns the minimum length required for dst. -func (f *freelist) copyall(dst []pgid) { - m := make(pgids, 0, f.pending_count()) - for _, list := range f.pending { - m = append(m, list...) - } - sort.Sort(m) - mergepgids(dst, f.ids, m) -} - -// allocate returns the starting page id of a contiguous list of pages of a given size. -// If a contiguous block cannot be found then 0 is returned. -func (f *freelist) allocate(n int) pgid { - if len(f.ids) == 0 { - return 0 - } - - var initial, previd pgid - for i, id := range f.ids { - if id <= 1 { - panic(fmt.Sprintf("invalid page allocation: %d", id)) - } - - // Reset initial page if this is not contiguous. - if previd == 0 || id-previd != 1 { - initial = id - } - - // If we found a contiguous block then remove it and return it. - if (id-initial)+1 == pgid(n) { - // If we're allocating off the beginning then take the fast path - // and just adjust the existing slice. This will use extra memory - // temporarily but the append() in free() will realloc the slice - // as is necessary. - if (i + 1) == n { - f.ids = f.ids[i+1:] - } else { - copy(f.ids[i-n+1:], f.ids[i+1:]) - f.ids = f.ids[:len(f.ids)-n] - } - - // Remove from the free cache. - for i := pgid(0); i < pgid(n); i++ { - delete(f.cache, initial+i) - } - - return initial - } - - previd = id - } - return 0 -} - -// free releases a page and its overflow for a given transaction id. -// If the page is already free then a panic will occur. -func (f *freelist) free(txid txid, p *page) { - if p.id <= 1 { - panic(fmt.Sprintf("cannot free page 0 or 1: %d", p.id)) - } - - // Free page and all its overflow pages. - var ids = f.pending[txid] - for id := p.id; id <= p.id+pgid(p.overflow); id++ { - // Verify that page is not already free. - if f.cache[id] { - panic(fmt.Sprintf("page %d already freed", id)) - } - - // Add to the freelist and cache. - ids = append(ids, id) - f.cache[id] = true - } - f.pending[txid] = ids -} - -// release moves all page ids for a transaction id (or older) to the freelist. -func (f *freelist) release(txid txid) { - m := make(pgids, 0) - for tid, ids := range f.pending { - if tid <= txid { - // Move transaction's pending pages to the available freelist. - // Don't remove from the cache since the page is still free. - m = append(m, ids...) - delete(f.pending, tid) - } - } - sort.Sort(m) - f.ids = pgids(f.ids).merge(m) -} - -// rollback removes the pages from a given pending tx. -func (f *freelist) rollback(txid txid) { - // Remove page ids from cache. - for _, id := range f.pending[txid] { - delete(f.cache, id) - } - - // Remove pages from pending list. - delete(f.pending, txid) -} - -// freed returns whether a given page is in the free list. -func (f *freelist) freed(pgid pgid) bool { - return f.cache[pgid] -} - -// read initializes the freelist from a freelist page. -func (f *freelist) read(p *page) { - // If the page.count is at the max uint16 value (64k) then it's considered - // an overflow and the size of the freelist is stored as the first element. - idx, count := 0, int(p.count) - if count == 0xFFFF { - idx = 1 - count = int(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[0]) - } - - // Copy the list of page ids from the freelist. - if count == 0 { - f.ids = nil - } else { - ids := ((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[idx:count] - f.ids = make([]pgid, len(ids)) - copy(f.ids, ids) - - // Make sure they're sorted. - sort.Sort(pgids(f.ids)) - } - - // Rebuild the page cache. - f.reindex() -} - -// write writes the page ids onto a freelist page. All free and pending ids are -// saved to disk since in the event of a program crash, all pending ids will -// become free. -func (f *freelist) write(p *page) error { - // Combine the old free pgids and pgids waiting on an open transaction. - - // Update the header flag. - p.flags |= freelistPageFlag - - // The page.count can only hold up to 64k elements so if we overflow that - // number then we handle it by putting the size in the first element. - lenids := f.count() - if lenids == 0 { - p.count = uint16(lenids) - } else if lenids < 0xFFFF { - p.count = uint16(lenids) - f.copyall(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[:]) - } else { - p.count = 0xFFFF - ((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[0] = pgid(lenids) - f.copyall(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[1:]) - } - - return nil -} - -// reload reads the freelist from a page and filters out pending items. -func (f *freelist) reload(p *page) { - f.read(p) - - // Build a cache of only pending pages. - pcache := make(map[pgid]bool) - for _, pendingIDs := range f.pending { - for _, pendingID := range pendingIDs { - pcache[pendingID] = true - } - } - - // Check each page in the freelist and build a new available freelist - // with any pages not in the pending lists. - var a []pgid - for _, id := range f.ids { - if !pcache[id] { - a = append(a, id) - } - } - f.ids = a - - // Once the available list is rebuilt then rebuild the free cache so that - // it includes the available and pending free pages. - f.reindex() -} - -// reindex rebuilds the free cache based on available and pending free lists. -func (f *freelist) reindex() { - f.cache = make(map[pgid]bool, len(f.ids)) - for _, id := range f.ids { - f.cache[id] = true - } - for _, pendingIDs := range f.pending { - for _, pendingID := range pendingIDs { - f.cache[pendingID] = true - } - } -} diff --git a/vendor/github.com/boltdb/bolt/.gitignore b/vendor/go.etcd.io/bbolt/.gitignore similarity index 72% rename from vendor/github.com/boltdb/bolt/.gitignore rename to vendor/go.etcd.io/bbolt/.gitignore index c7bd2b7a..3bcd8cba 100644 --- a/vendor/github.com/boltdb/bolt/.gitignore +++ b/vendor/go.etcd.io/bbolt/.gitignore @@ -2,3 +2,4 @@ *.test *.swp /bin/ +cover.out diff --git a/vendor/go.etcd.io/bbolt/.travis.yml b/vendor/go.etcd.io/bbolt/.travis.yml new file mode 100644 index 00000000..257dfdfe --- /dev/null +++ b/vendor/go.etcd.io/bbolt/.travis.yml @@ -0,0 +1,17 @@ +language: go +go_import_path: go.etcd.io/bbolt + +sudo: false + +go: +- 1.12 + +before_install: +- go get -v honnef.co/go/tools/... +- go get -v github.com/kisielk/errcheck + +script: +- make fmt +- make test +- make race +# - make errcheck diff --git a/vendor/github.com/boltdb/bolt/LICENSE b/vendor/go.etcd.io/bbolt/LICENSE similarity index 100% rename from vendor/github.com/boltdb/bolt/LICENSE rename to vendor/go.etcd.io/bbolt/LICENSE diff --git a/vendor/go.etcd.io/bbolt/Makefile b/vendor/go.etcd.io/bbolt/Makefile new file mode 100644 index 00000000..2968aaa6 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/Makefile @@ -0,0 +1,38 @@ +BRANCH=`git rev-parse --abbrev-ref HEAD` +COMMIT=`git rev-parse --short HEAD` +GOLDFLAGS="-X main.branch $(BRANCH) -X main.commit $(COMMIT)" + +default: build + +race: + @TEST_FREELIST_TYPE=hashmap go test -v -race -test.run="TestSimulate_(100op|1000op)" + @echo "array freelist test" + @TEST_FREELIST_TYPE=array go test -v -race -test.run="TestSimulate_(100op|1000op)" + +fmt: + !(gofmt -l -s -d $(shell find . -name \*.go) | grep '[a-z]') + +# go get honnef.co/go/tools/simple +gosimple: + gosimple ./... + +# go get honnef.co/go/tools/unused +unused: + unused ./... + +# go get github.com/kisielk/errcheck +errcheck: + @errcheck -ignorepkg=bytes -ignore=os:Remove go.etcd.io/bbolt + +test: + TEST_FREELIST_TYPE=hashmap go test -timeout 20m -v -coverprofile cover.out -covermode atomic + # Note: gets "program not an importable package" in out of path builds + TEST_FREELIST_TYPE=hashmap go test -v ./cmd/bbolt + + @echo "array freelist test" + + @TEST_FREELIST_TYPE=array go test -timeout 20m -v -coverprofile cover.out -covermode atomic + # Note: gets "program not an importable package" in out of path builds + @TEST_FREELIST_TYPE=array go test -v ./cmd/bbolt + +.PHONY: race fmt errcheck test gosimple unused diff --git a/vendor/github.com/boltdb/bolt/README.md b/vendor/go.etcd.io/bbolt/README.md similarity index 87% rename from vendor/github.com/boltdb/bolt/README.md rename to vendor/go.etcd.io/bbolt/README.md index 7d43a15b..c9e64b1a 100644 --- a/vendor/github.com/boltdb/bolt/README.md +++ b/vendor/go.etcd.io/bbolt/README.md @@ -1,5 +1,18 @@ -Bolt [![Coverage Status](https://coveralls.io/repos/boltdb/bolt/badge.svg?branch=master)](https://coveralls.io/r/boltdb/bolt?branch=master) [![GoDoc](https://godoc.org/github.com/boltdb/bolt?status.svg)](https://godoc.org/github.com/boltdb/bolt) ![Version](https://img.shields.io/badge/version-1.2.1-green.svg) -==== +bbolt +===== + +[![Go Report Card](https://goreportcard.com/badge/github.com/etcd-io/bbolt?style=flat-square)](https://goreportcard.com/report/github.com/etcd-io/bbolt) +[![Coverage](https://codecov.io/gh/etcd-io/bbolt/branch/master/graph/badge.svg)](https://codecov.io/gh/etcd-io/bbolt) +[![Build Status Travis](https://img.shields.io/travis/etcd-io/bboltlabs.svg?style=flat-square&&branch=master)](https://travis-ci.com/etcd-io/bbolt) +[![Godoc](http://img.shields.io/badge/go-documentation-blue.svg?style=flat-square)](https://godoc.org/github.com/etcd-io/bbolt) +[![Releases](https://img.shields.io/github/release/etcd-io/bbolt/all.svg?style=flat-square)](https://github.com/etcd-io/bbolt/releases) +[![LICENSE](https://img.shields.io/github/license/etcd-io/bbolt.svg?style=flat-square)](https://github.com/etcd-io/bbolt/blob/master/LICENSE) + +bbolt is a fork of [Ben Johnson's][gh_ben] [Bolt][bolt] key/value +store. The purpose of this fork is to provide the Go community with an active +maintenance and development target for Bolt; the goal is improved reliability +and stability. bbolt includes bug fixes, performance enhancements, and features +not found in Bolt while preserving backwards compatibility with the Bolt API. Bolt is a pure Go key/value store inspired by [Howard Chu's][hyc_symas] [LMDB project][lmdb]. The goal of the project is to provide a simple, @@ -10,6 +23,8 @@ Since Bolt is meant to be used as such a low-level piece of functionality, simplicity is key. The API will be small and only focus on getting values and setting values. That's it. +[gh_ben]: https://github.com/benbjohnson +[bolt]: https://github.com/boltdb/bolt [hyc_symas]: https://twitter.com/hyc_symas [lmdb]: http://symas.com/mdb/ @@ -21,36 +36,42 @@ consistency and thread safety. Bolt is currently used in high-load production environments serving databases as large as 1TB. Many companies such as Shopify and Heroku use Bolt-backed services every day. +## Project versioning + +bbolt uses [semantic versioning](http://semver.org). +API should not change between patch and minor releases. +New minor versions may add additional features to the API. + ## Table of Contents -- [Getting Started](#getting-started) - - [Installing](#installing) - - [Opening a database](#opening-a-database) - - [Transactions](#transactions) - - [Read-write transactions](#read-write-transactions) - - [Read-only transactions](#read-only-transactions) - - [Batch read-write transactions](#batch-read-write-transactions) - - [Managing transactions manually](#managing-transactions-manually) - - [Using buckets](#using-buckets) - - [Using key/value pairs](#using-keyvalue-pairs) - - [Autoincrementing integer for the bucket](#autoincrementing-integer-for-the-bucket) - - [Iterating over keys](#iterating-over-keys) - - [Prefix scans](#prefix-scans) - - [Range scans](#range-scans) - - [ForEach()](#foreach) - - [Nested buckets](#nested-buckets) - - [Database backups](#database-backups) - - [Statistics](#statistics) - - [Read-Only Mode](#read-only-mode) - - [Mobile Use (iOS/Android)](#mobile-use-iosandroid) -- [Resources](#resources) -- [Comparison with other databases](#comparison-with-other-databases) - - [Postgres, MySQL, & other relational databases](#postgres-mysql--other-relational-databases) - - [LevelDB, RocksDB](#leveldb-rocksdb) - - [LMDB](#lmdb) -- [Caveats & Limitations](#caveats--limitations) -- [Reading the Source](#reading-the-source) -- [Other Projects Using Bolt](#other-projects-using-bolt) + - [Getting Started](#getting-started) + - [Installing](#installing) + - [Opening a database](#opening-a-database) + - [Transactions](#transactions) + - [Read-write transactions](#read-write-transactions) + - [Read-only transactions](#read-only-transactions) + - [Batch read-write transactions](#batch-read-write-transactions) + - [Managing transactions manually](#managing-transactions-manually) + - [Using buckets](#using-buckets) + - [Using key/value pairs](#using-keyvalue-pairs) + - [Autoincrementing integer for the bucket](#autoincrementing-integer-for-the-bucket) + - [Iterating over keys](#iterating-over-keys) + - [Prefix scans](#prefix-scans) + - [Range scans](#range-scans) + - [ForEach()](#foreach) + - [Nested buckets](#nested-buckets) + - [Database backups](#database-backups) + - [Statistics](#statistics) + - [Read-Only Mode](#read-only-mode) + - [Mobile Use (iOS/Android)](#mobile-use-iosandroid) + - [Resources](#resources) + - [Comparison with other databases](#comparison-with-other-databases) + - [Postgres, MySQL, & other relational databases](#postgres-mysql--other-relational-databases) + - [LevelDB, RocksDB](#leveldb-rocksdb) + - [LMDB](#lmdb) + - [Caveats & Limitations](#caveats--limitations) + - [Reading the Source](#reading-the-source) + - [Other Projects Using Bolt](#other-projects-using-bolt) ## Getting Started @@ -59,13 +80,28 @@ Shopify and Heroku use Bolt-backed services every day. To start using Bolt, install Go and run `go get`: ```sh -$ go get github.com/boltdb/bolt/... +$ go get go.etcd.io/bbolt/... ``` This will retrieve the library and install the `bolt` command line utility into your `$GOBIN` path. +### Importing bbolt + +To use bbolt as an embedded key-value store, import as: + +```go +import bolt "go.etcd.io/bbolt" + +db, err := bolt.Open(path, 0666, nil) +if err != nil { + return err +} +defer db.Close() +``` + + ### Opening a database The top-level object in Bolt is a `DB`. It is represented as a single file on @@ -79,7 +115,7 @@ package main import ( "log" - "github.com/boltdb/bolt" + bolt "go.etcd.io/bbolt" ) func main() { @@ -116,11 +152,12 @@ are not thread safe. To work with data in multiple goroutines you must start a transaction for each one or use locking to ensure only one goroutine accesses a transaction at a time. Creating transaction from the `DB` is thread safe. -Read-only transactions and read-write transactions should not depend on one -another and generally shouldn't be opened simultaneously in the same goroutine. -This can cause a deadlock as the read-write transaction needs to periodically -re-map the data file but it cannot do so while a read-only transaction is open. - +Transactions should not depend on one another and generally shouldn't be opened +simultaneously in the same goroutine. This can cause a deadlock as the read-write +transaction needs to periodically re-map the data file but it cannot do so while +any read-only transaction is open. Even a nested read-only transaction can cause +a deadlock, as the child transaction can block the parent transaction from releasing +its resources. #### Read-write transactions @@ -239,7 +276,7 @@ should be writable. ### Using buckets Buckets are collections of key/value pairs within the database. All keys in a -bucket must be unique. You can create a bucket using the `DB.CreateBucket()` +bucket must be unique. You can create a bucket using the `Tx.CreateBucket()` function: ```go @@ -522,7 +559,7 @@ this from a read-only transaction, it will perform a hot backup and not block your other database reads and writes. By default, it will use a regular file handle which will utilize the operating -system's page cache. See the [`Tx`](https://godoc.org/github.com/boltdb/bolt#Tx) +system's page cache. See the [`Tx`](https://godoc.org/go.etcd.io/bbolt#Tx) documentation for information about optimizing for larger-than-RAM datasets. One common use case is to backup over HTTP so you can use tools like `cURL` to @@ -811,7 +848,7 @@ Here are a few things to note when evaluating and using Bolt: ## Reading the Source -Bolt is a relatively small code base (<3KLOC) for an embedded, serializable, +Bolt is a relatively small code base (<5KLOC) for an embedded, serializable, transactional key/value database so it can be a good starting point for people interested in how databases work. @@ -863,54 +900,58 @@ them via pull request. Below is a list of public, open source projects that use Bolt: -* [BoltDbWeb](https://github.com/evnix/boltdbweb) - A web based GUI for BoltDB files. -* [Operation Go: A Routine Mission](http://gocode.io) - An online programming game for Golang using Bolt for user accounts and a leaderboard. +* [Algernon](https://github.com/xyproto/algernon) - A HTTP/2 web server with built-in support for Lua. Uses BoltDB as the default database backend. * [Bazil](https://bazil.org/) - A file system that lets your data reside where it is most convenient for it to reside. -* [DVID](https://github.com/janelia-flyem/dvid) - Added Bolt as optional storage engine and testing it against Basho-tuned leveldb. -* [Skybox Analytics](https://github.com/skybox/skybox) - A standalone funnel analysis tool for web analytics. -* [Scuttlebutt](https://github.com/benbjohnson/scuttlebutt) - Uses Bolt to store and process all Twitter mentions of GitHub projects. -* [Wiki](https://github.com/peterhellberg/wiki) - A tiny wiki using Goji, BoltDB and Blackfriday. -* [ChainStore](https://github.com/pressly/chainstore) - Simple key-value interface to a variety of storage engines organized as a chain of operations. -* [MetricBase](https://github.com/msiebuhr/MetricBase) - Single-binary version of Graphite. -* [Gitchain](https://github.com/gitchain/gitchain) - Decentralized, peer-to-peer Git repositories aka "Git meets Bitcoin". -* [event-shuttle](https://github.com/sclasen/event-shuttle) - A Unix system service to collect and reliably deliver messages to Kafka. -* [ipxed](https://github.com/kelseyhightower/ipxed) - Web interface and api for ipxed. +* [bolter](https://github.com/hasit/bolter) - Command-line app for viewing BoltDB file in your terminal. +* [boltcli](https://github.com/spacewander/boltcli) - the redis-cli for boltdb with Lua script support. +* [BoltHold](https://github.com/timshannon/bolthold) - An embeddable NoSQL store for Go types built on BoltDB * [BoltStore](https://github.com/yosssi/boltstore) - Session store using Bolt. -* [photosite/session](https://godoc.org/bitbucket.org/kardianos/photosite/session) - Sessions for a photo viewing site. -* [LedisDB](https://github.com/siddontang/ledisdb) - A high performance NoSQL, using Bolt as optional storage. -* [ipLocator](https://github.com/AndreasBriese/ipLocator) - A fast ip-geo-location-server using bolt with bloom filters. -* [cayley](https://github.com/google/cayley) - Cayley is an open-source graph database using Bolt as optional backend. +* [Boltdb Boilerplate](https://github.com/bobintornado/boltdb-boilerplate) - Boilerplate wrapper around bolt aiming to make simple calls one-liners. +* [BoltDbWeb](https://github.com/evnix/boltdbweb) - A web based GUI for BoltDB files. * [bleve](http://www.blevesearch.com/) - A pure Go search engine similar to ElasticSearch that uses Bolt as the default storage backend. -* [tentacool](https://github.com/optiflows/tentacool) - REST api server to manage system stuff (IP, DNS, Gateway...) on a linux server. -* [Seaweed File System](https://github.com/chrislusf/seaweedfs) - Highly scalable distributed key~file system with O(1) disk read. -* [InfluxDB](https://influxdata.com) - Scalable datastore for metrics, events, and real-time analytics. -* [Freehold](http://tshannon.bitbucket.org/freehold/) - An open, secure, and lightweight platform for your files and data. -* [Prometheus Annotation Server](https://github.com/oliver006/prom_annotation_server) - Annotation server for PromDash & Prometheus service monitoring system. -* [Consul](https://github.com/hashicorp/consul) - Consul is service discovery and configuration made easy. Distributed, highly available, and datacenter-aware. -* [Kala](https://github.com/ajvb/kala) - Kala is a modern job scheduler optimized to run on a single node. It is persistent, JSON over HTTP API, ISO 8601 duration notation, and dependent jobs. -* [drive](https://github.com/odeke-em/drive) - drive is an unofficial Google Drive command line client for \*NIX operating systems. -* [stow](https://github.com/djherbis/stow) - a persistence manager for objects - backed by boltdb. +* [btcwallet](https://github.com/btcsuite/btcwallet) - A bitcoin wallet. * [buckets](https://github.com/joyrexus/buckets) - a bolt wrapper streamlining simple tx and key scans. -* [mbuckets](https://github.com/abhigupta912/mbuckets) - A Bolt wrapper that allows easy operations on multi level (nested) buckets. -* [Request Baskets](https://github.com/darklynx/request-baskets) - A web service to collect arbitrary HTTP requests and inspect them via REST API or simple web UI, similar to [RequestBin](http://requestb.in/) service -* [Go Report Card](https://goreportcard.com/) - Go code quality report cards as a (free and open source) service. -* [Boltdb Boilerplate](https://github.com/bobintornado/boltdb-boilerplate) - Boilerplate wrapper around bolt aiming to make simple calls one-liners. -* [lru](https://github.com/crowdriff/lru) - Easy to use Bolt-backed Least-Recently-Used (LRU) read-through cache with chainable remote stores. -* [Storm](https://github.com/asdine/storm) - Simple and powerful ORM for BoltDB. -* [GoWebApp](https://github.com/josephspurrier/gowebapp) - A basic MVC web application in Go using BoltDB. -* [SimpleBolt](https://github.com/xyproto/simplebolt) - A simple way to use BoltDB. Deals mainly with strings. -* [Algernon](https://github.com/xyproto/algernon) - A HTTP/2 web server with built-in support for Lua. Uses BoltDB as the default database backend. -* [MuLiFS](https://github.com/dankomiocevic/mulifs) - Music Library Filesystem creates a filesystem to organise your music files. -* [GoShort](https://github.com/pankajkhairnar/goShort) - GoShort is a URL shortener written in Golang and BoltDB for persistent key/value storage and for routing it's using high performent HTTPRouter. -* [torrent](https://github.com/anacrolix/torrent) - Full-featured BitTorrent client package and utilities in Go. BoltDB is a storage backend in development. -* [gopherpit](https://github.com/gopherpit/gopherpit) - A web service to manage Go remote import paths with custom domains -* [bolter](https://github.com/hasit/bolter) - Command-line app for viewing BoltDB file in your terminal. -* [btcwallet](https://github.com/btcsuite/btcwallet) - A bitcoin wallet. +* [cayley](https://github.com/google/cayley) - Cayley is an open-source graph database using Bolt as optional backend. +* [ChainStore](https://github.com/pressly/chainstore) - Simple key-value interface to a variety of storage engines organized as a chain of operations. +* [Consul](https://github.com/hashicorp/consul) - Consul is service discovery and configuration made easy. Distributed, highly available, and datacenter-aware. +* [DVID](https://github.com/janelia-flyem/dvid) - Added Bolt as optional storage engine and testing it against Basho-tuned leveldb. * [dcrwallet](https://github.com/decred/dcrwallet) - A wallet for the Decred cryptocurrency. +* [drive](https://github.com/odeke-em/drive) - drive is an unofficial Google Drive command line client for \*NIX operating systems. +* [event-shuttle](https://github.com/sclasen/event-shuttle) - A Unix system service to collect and reliably deliver messages to Kafka. +* [Freehold](http://tshannon.bitbucket.org/freehold/) - An open, secure, and lightweight platform for your files and data. +* [Go Report Card](https://goreportcard.com/) - Go code quality report cards as a (free and open source) service. +* [GoWebApp](https://github.com/josephspurrier/gowebapp) - A basic MVC web application in Go using BoltDB. +* [GoShort](https://github.com/pankajkhairnar/goShort) - GoShort is a URL shortener written in Golang and BoltDB for persistent key/value storage and for routing it's using high performent HTTPRouter. +* [gopherpit](https://github.com/gopherpit/gopherpit) - A web service to manage Go remote import paths with custom domains +* [gokv](https://github.com/philippgille/gokv) - Simple key-value store abstraction and implementations for Go (Redis, Consul, etcd, bbolt, BadgerDB, LevelDB, Memcached, DynamoDB, S3, PostgreSQL, MongoDB, CockroachDB and many more) +* [Gitchain](https://github.com/gitchain/gitchain) - Decentralized, peer-to-peer Git repositories aka "Git meets Bitcoin". +* [InfluxDB](https://influxdata.com) - Scalable datastore for metrics, events, and real-time analytics. +* [ipLocator](https://github.com/AndreasBriese/ipLocator) - A fast ip-geo-location-server using bolt with bloom filters. +* [ipxed](https://github.com/kelseyhightower/ipxed) - Web interface and api for ipxed. * [Ironsmith](https://github.com/timshannon/ironsmith) - A simple, script-driven continuous integration (build - > test -> release) tool, with no external dependencies -* [BoltHold](https://github.com/timshannon/bolthold) - An embeddable NoSQL store for Go types built on BoltDB -* [Ponzu CMS](https://ponzu-cms.org) - Headless CMS + automatic JSON API with auto-HTTPS, HTTP/2 Server Push, and flexible server framework. +* [Kala](https://github.com/ajvb/kala) - Kala is a modern job scheduler optimized to run on a single node. It is persistent, JSON over HTTP API, ISO 8601 duration notation, and dependent jobs. +* [Key Value Access Langusge (KVAL)](https://github.com/kval-access-language) - A proposed grammar for key-value datastores offering a bbolt binding. +* [LedisDB](https://github.com/siddontang/ledisdb) - A high performance NoSQL, using Bolt as optional storage. +* [lru](https://github.com/crowdriff/lru) - Easy to use Bolt-backed Least-Recently-Used (LRU) read-through cache with chainable remote stores. +* [mbuckets](https://github.com/abhigupta912/mbuckets) - A Bolt wrapper that allows easy operations on multi level (nested) buckets. +* [MetricBase](https://github.com/msiebuhr/MetricBase) - Single-binary version of Graphite. +* [MuLiFS](https://github.com/dankomiocevic/mulifs) - Music Library Filesystem creates a filesystem to organise your music files. +* [NATS](https://github.com/nats-io/nats-streaming-server) - NATS Streaming uses bbolt for message and metadata storage. +* [Operation Go: A Routine Mission](http://gocode.io) - An online programming game for Golang using Bolt for user accounts and a leaderboard. +* [photosite/session](https://godoc.org/bitbucket.org/kardianos/photosite/session) - Sessions for a photo viewing site. +* [Prometheus Annotation Server](https://github.com/oliver006/prom_annotation_server) - Annotation server for PromDash & Prometheus service monitoring system. +* [reef-pi](https://github.com/reef-pi/reef-pi) - reef-pi is an award winning, modular, DIY reef tank controller using easy to learn electronics based on a Raspberry Pi. +* [Request Baskets](https://github.com/darklynx/request-baskets) - A web service to collect arbitrary HTTP requests and inspect them via REST API or simple web UI, similar to [RequestBin](http://requestb.in/) service +* [Seaweed File System](https://github.com/chrislusf/seaweedfs) - Highly scalable distributed key~file system with O(1) disk read. +* [stow](https://github.com/djherbis/stow) - a persistence manager for objects + backed by boltdb. +* [Storm](https://github.com/asdine/storm) - Simple and powerful ORM for BoltDB. +* [SimpleBolt](https://github.com/xyproto/simplebolt) - A simple way to use BoltDB. Deals mainly with strings. +* [Skybox Analytics](https://github.com/skybox/skybox) - A standalone funnel analysis tool for web analytics. +* [Scuttlebutt](https://github.com/benbjohnson/scuttlebutt) - Uses Bolt to store and process all Twitter mentions of GitHub projects. +* [tentacool](https://github.com/optiflows/tentacool) - REST api server to manage system stuff (IP, DNS, Gateway...) on a linux server. +* [torrent](https://github.com/anacrolix/torrent) - Full-featured BitTorrent client package and utilities in Go. BoltDB is a storage backend in development. +* [Wiki](https://github.com/peterhellberg/wiki) - A tiny wiki using Goji, BoltDB and Blackfriday. If you are using Bolt in a project please send a pull request to add it to the list. diff --git a/vendor/github.com/boltdb/bolt/bolt_386.go b/vendor/go.etcd.io/bbolt/bolt_386.go similarity index 68% rename from vendor/github.com/boltdb/bolt/bolt_386.go rename to vendor/go.etcd.io/bbolt/bolt_386.go index 820d533c..aee25960 100644 --- a/vendor/github.com/boltdb/bolt/bolt_386.go +++ b/vendor/go.etcd.io/bbolt/bolt_386.go @@ -1,10 +1,7 @@ -package bolt +package bbolt // maxMapSize represents the largest mmap size supported by Bolt. const maxMapSize = 0x7FFFFFFF // 2GB // maxAllocSize is the size used when creating array pointers. const maxAllocSize = 0xFFFFFFF - -// Are unaligned load/stores broken on this arch? -var brokenUnaligned = false diff --git a/vendor/go.etcd.io/bbolt/bolt_amd64.go b/vendor/go.etcd.io/bbolt/bolt_amd64.go new file mode 100644 index 00000000..5dd8f3f2 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/bolt_amd64.go @@ -0,0 +1,7 @@ +package bbolt + +// maxMapSize represents the largest mmap size supported by Bolt. +const maxMapSize = 0xFFFFFFFFFFFF // 256TB + +// maxAllocSize is the size used when creating array pointers. +const maxAllocSize = 0x7FFFFFFF diff --git a/vendor/go.etcd.io/bbolt/bolt_arm.go b/vendor/go.etcd.io/bbolt/bolt_arm.go new file mode 100644 index 00000000..aee25960 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/bolt_arm.go @@ -0,0 +1,7 @@ +package bbolt + +// maxMapSize represents the largest mmap size supported by Bolt. +const maxMapSize = 0x7FFFFFFF // 2GB + +// maxAllocSize is the size used when creating array pointers. +const maxAllocSize = 0xFFFFFFF diff --git a/vendor/github.com/boltdb/bolt/bolt_arm64.go b/vendor/go.etcd.io/bbolt/bolt_arm64.go similarity index 70% rename from vendor/github.com/boltdb/bolt/bolt_arm64.go rename to vendor/go.etcd.io/bbolt/bolt_arm64.go index b26d84f9..810dfd55 100644 --- a/vendor/github.com/boltdb/bolt/bolt_arm64.go +++ b/vendor/go.etcd.io/bbolt/bolt_arm64.go @@ -1,12 +1,9 @@ // +build arm64 -package bolt +package bbolt // maxMapSize represents the largest mmap size supported by Bolt. const maxMapSize = 0xFFFFFFFFFFFF // 256TB // maxAllocSize is the size used when creating array pointers. const maxAllocSize = 0x7FFFFFFF - -// Are unaligned load/stores broken on this arch? -var brokenUnaligned = false diff --git a/vendor/github.com/boltdb/bolt/bolt_linux.go b/vendor/go.etcd.io/bbolt/bolt_linux.go similarity index 91% rename from vendor/github.com/boltdb/bolt/bolt_linux.go rename to vendor/go.etcd.io/bbolt/bolt_linux.go index 2b676661..7707bcac 100644 --- a/vendor/github.com/boltdb/bolt/bolt_linux.go +++ b/vendor/go.etcd.io/bbolt/bolt_linux.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "syscall" diff --git a/vendor/go.etcd.io/bbolt/bolt_mips64x.go b/vendor/go.etcd.io/bbolt/bolt_mips64x.go new file mode 100644 index 00000000..dd8ffe12 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/bolt_mips64x.go @@ -0,0 +1,9 @@ +// +build mips64 mips64le + +package bbolt + +// maxMapSize represents the largest mmap size supported by Bolt. +const maxMapSize = 0x8000000000 // 512GB + +// maxAllocSize is the size used when creating array pointers. +const maxAllocSize = 0x7FFFFFFF diff --git a/vendor/go.etcd.io/bbolt/bolt_mipsx.go b/vendor/go.etcd.io/bbolt/bolt_mipsx.go new file mode 100644 index 00000000..a669703a --- /dev/null +++ b/vendor/go.etcd.io/bbolt/bolt_mipsx.go @@ -0,0 +1,9 @@ +// +build mips mipsle + +package bbolt + +// maxMapSize represents the largest mmap size supported by Bolt. +const maxMapSize = 0x40000000 // 1GB + +// maxAllocSize is the size used when creating array pointers. +const maxAllocSize = 0xFFFFFFF diff --git a/vendor/github.com/boltdb/bolt/bolt_openbsd.go b/vendor/go.etcd.io/bbolt/bolt_openbsd.go similarity index 97% rename from vendor/github.com/boltdb/bolt/bolt_openbsd.go rename to vendor/go.etcd.io/bbolt/bolt_openbsd.go index 7058c3d7..d7f50358 100644 --- a/vendor/github.com/boltdb/bolt/bolt_openbsd.go +++ b/vendor/go.etcd.io/bbolt/bolt_openbsd.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "syscall" diff --git a/vendor/github.com/boltdb/bolt/bolt_ppc.go b/vendor/go.etcd.io/bbolt/bolt_ppc.go similarity index 93% rename from vendor/github.com/boltdb/bolt/bolt_ppc.go rename to vendor/go.etcd.io/bbolt/bolt_ppc.go index 645ddc3e..84e545ef 100644 --- a/vendor/github.com/boltdb/bolt/bolt_ppc.go +++ b/vendor/go.etcd.io/bbolt/bolt_ppc.go @@ -1,6 +1,6 @@ // +build ppc -package bolt +package bbolt // maxMapSize represents the largest mmap size supported by Bolt. const maxMapSize = 0x7FFFFFFF // 2GB diff --git a/vendor/github.com/boltdb/bolt/bolt_ppc64.go b/vendor/go.etcd.io/bbolt/bolt_ppc64.go similarity index 70% rename from vendor/github.com/boltdb/bolt/bolt_ppc64.go rename to vendor/go.etcd.io/bbolt/bolt_ppc64.go index 9331d977..a7612090 100644 --- a/vendor/github.com/boltdb/bolt/bolt_ppc64.go +++ b/vendor/go.etcd.io/bbolt/bolt_ppc64.go @@ -1,12 +1,9 @@ // +build ppc64 -package bolt +package bbolt // maxMapSize represents the largest mmap size supported by Bolt. const maxMapSize = 0xFFFFFFFFFFFF // 256TB // maxAllocSize is the size used when creating array pointers. const maxAllocSize = 0x7FFFFFFF - -// Are unaligned load/stores broken on this arch? -var brokenUnaligned = false diff --git a/vendor/github.com/boltdb/bolt/bolt_ppc64le.go b/vendor/go.etcd.io/bbolt/bolt_ppc64le.go similarity index 70% rename from vendor/github.com/boltdb/bolt/bolt_ppc64le.go rename to vendor/go.etcd.io/bbolt/bolt_ppc64le.go index 8c143bc5..c830f2fc 100644 --- a/vendor/github.com/boltdb/bolt/bolt_ppc64le.go +++ b/vendor/go.etcd.io/bbolt/bolt_ppc64le.go @@ -1,12 +1,9 @@ // +build ppc64le -package bolt +package bbolt // maxMapSize represents the largest mmap size supported by Bolt. const maxMapSize = 0xFFFFFFFFFFFF // 256TB // maxAllocSize is the size used when creating array pointers. const maxAllocSize = 0x7FFFFFFF - -// Are unaligned load/stores broken on this arch? -var brokenUnaligned = false diff --git a/vendor/github.com/boltdb/bolt/bolt_amd64.go b/vendor/go.etcd.io/bbolt/bolt_riscv64.go similarity index 69% rename from vendor/github.com/boltdb/bolt/bolt_amd64.go rename to vendor/go.etcd.io/bbolt/bolt_riscv64.go index 98fafdb4..c967613b 100644 --- a/vendor/github.com/boltdb/bolt/bolt_amd64.go +++ b/vendor/go.etcd.io/bbolt/bolt_riscv64.go @@ -1,10 +1,9 @@ -package bolt +// +build riscv64 + +package bbolt // maxMapSize represents the largest mmap size supported by Bolt. const maxMapSize = 0xFFFFFFFFFFFF // 256TB // maxAllocSize is the size used when creating array pointers. const maxAllocSize = 0x7FFFFFFF - -// Are unaligned load/stores broken on this arch? -var brokenUnaligned = false diff --git a/vendor/github.com/boltdb/bolt/bolt_s390x.go b/vendor/go.etcd.io/bbolt/bolt_s390x.go similarity index 70% rename from vendor/github.com/boltdb/bolt/bolt_s390x.go rename to vendor/go.etcd.io/bbolt/bolt_s390x.go index d7c39af9..ff2a5609 100644 --- a/vendor/github.com/boltdb/bolt/bolt_s390x.go +++ b/vendor/go.etcd.io/bbolt/bolt_s390x.go @@ -1,12 +1,9 @@ // +build s390x -package bolt +package bbolt // maxMapSize represents the largest mmap size supported by Bolt. const maxMapSize = 0xFFFFFFFFFFFF // 256TB // maxAllocSize is the size used when creating array pointers. const maxAllocSize = 0x7FFFFFFF - -// Are unaligned load/stores broken on this arch? -var brokenUnaligned = false diff --git a/vendor/github.com/boltdb/bolt/bolt_unix.go b/vendor/go.etcd.io/bbolt/bolt_unix.go similarity index 70% rename from vendor/github.com/boltdb/bolt/bolt_unix.go rename to vendor/go.etcd.io/bbolt/bolt_unix.go index cad62dda..2938fed5 100644 --- a/vendor/github.com/boltdb/bolt/bolt_unix.go +++ b/vendor/go.etcd.io/bbolt/bolt_unix.go @@ -1,41 +1,43 @@ -// +build !windows,!plan9,!solaris +// +build !windows,!plan9,!solaris,!aix -package bolt +package bbolt import ( "fmt" - "os" "syscall" "time" "unsafe" ) // flock acquires an advisory lock on a file descriptor. -func flock(db *DB, mode os.FileMode, exclusive bool, timeout time.Duration) error { +func flock(db *DB, exclusive bool, timeout time.Duration) error { var t time.Time + if timeout != 0 { + t = time.Now() + } + fd := db.file.Fd() + flag := syscall.LOCK_NB + if exclusive { + flag |= syscall.LOCK_EX + } else { + flag |= syscall.LOCK_SH + } for { - // If we're beyond our timeout then return an error. - // This can only occur after we've attempted a flock once. - if t.IsZero() { - t = time.Now() - } else if timeout > 0 && time.Since(t) > timeout { - return ErrTimeout - } - flag := syscall.LOCK_SH - if exclusive { - flag = syscall.LOCK_EX - } - - // Otherwise attempt to obtain an exclusive lock. - err := syscall.Flock(int(db.file.Fd()), flag|syscall.LOCK_NB) + // Attempt to obtain an exclusive lock. + err := syscall.Flock(int(fd), flag) if err == nil { return nil } else if err != syscall.EWOULDBLOCK { return err } + // If we timed out then return an error. + if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout { + return ErrTimeout + } + // Wait for a bit and try again. - time.Sleep(50 * time.Millisecond) + time.Sleep(flockRetryTimeout) } } @@ -53,7 +55,9 @@ func mmap(db *DB, sz int) error { } // Advise the kernel that the mmap is accessed randomly. - if err := madvise(b, syscall.MADV_RANDOM); err != nil { + err = madvise(b, syscall.MADV_RANDOM) + if err != nil && err != syscall.ENOSYS { + // Ignore not implemented error in kernel because it still works. return fmt.Errorf("madvise: %s", err) } diff --git a/vendor/go.etcd.io/bbolt/bolt_unix_aix.go b/vendor/go.etcd.io/bbolt/bolt_unix_aix.go new file mode 100644 index 00000000..a64c16f5 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/bolt_unix_aix.go @@ -0,0 +1,90 @@ +// +build aix + +package bbolt + +import ( + "fmt" + "syscall" + "time" + "unsafe" + + "golang.org/x/sys/unix" +) + +// flock acquires an advisory lock on a file descriptor. +func flock(db *DB, exclusive bool, timeout time.Duration) error { + var t time.Time + if timeout != 0 { + t = time.Now() + } + fd := db.file.Fd() + var lockType int16 + if exclusive { + lockType = syscall.F_WRLCK + } else { + lockType = syscall.F_RDLCK + } + for { + // Attempt to obtain an exclusive lock. + lock := syscall.Flock_t{Type: lockType} + err := syscall.FcntlFlock(fd, syscall.F_SETLK, &lock) + if err == nil { + return nil + } else if err != syscall.EAGAIN { + return err + } + + // If we timed out then return an error. + if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout { + return ErrTimeout + } + + // Wait for a bit and try again. + time.Sleep(flockRetryTimeout) + } +} + +// funlock releases an advisory lock on a file descriptor. +func funlock(db *DB) error { + var lock syscall.Flock_t + lock.Start = 0 + lock.Len = 0 + lock.Type = syscall.F_UNLCK + lock.Whence = 0 + return syscall.FcntlFlock(uintptr(db.file.Fd()), syscall.F_SETLK, &lock) +} + +// mmap memory maps a DB's data file. +func mmap(db *DB, sz int) error { + // Map the data file to memory. + b, err := unix.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags) + if err != nil { + return err + } + + // Advise the kernel that the mmap is accessed randomly. + if err := unix.Madvise(b, syscall.MADV_RANDOM); err != nil { + return fmt.Errorf("madvise: %s", err) + } + + // Save the original byte slice and convert to a byte array pointer. + db.dataref = b + db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0])) + db.datasz = sz + return nil +} + +// munmap unmaps a DB's data file from memory. +func munmap(db *DB) error { + // Ignore the unmap if we have no mapped data. + if db.dataref == nil { + return nil + } + + // Unmap using the original byte slice. + err := unix.Munmap(db.dataref) + db.dataref = nil + db.data = nil + db.datasz = 0 + return err +} diff --git a/vendor/github.com/boltdb/bolt/bolt_unix_solaris.go b/vendor/go.etcd.io/bbolt/bolt_unix_solaris.go similarity index 70% rename from vendor/github.com/boltdb/bolt/bolt_unix_solaris.go rename to vendor/go.etcd.io/bbolt/bolt_unix_solaris.go index 307bf2b3..babad657 100644 --- a/vendor/github.com/boltdb/bolt/bolt_unix_solaris.go +++ b/vendor/go.etcd.io/bbolt/bolt_unix_solaris.go @@ -1,8 +1,7 @@ -package bolt +package bbolt import ( "fmt" - "os" "syscall" "time" "unsafe" @@ -11,36 +10,35 @@ import ( ) // flock acquires an advisory lock on a file descriptor. -func flock(db *DB, mode os.FileMode, exclusive bool, timeout time.Duration) error { +func flock(db *DB, exclusive bool, timeout time.Duration) error { var t time.Time + if timeout != 0 { + t = time.Now() + } + fd := db.file.Fd() + var lockType int16 + if exclusive { + lockType = syscall.F_WRLCK + } else { + lockType = syscall.F_RDLCK + } for { - // If we're beyond our timeout then return an error. - // This can only occur after we've attempted a flock once. - if t.IsZero() { - t = time.Now() - } else if timeout > 0 && time.Since(t) > timeout { - return ErrTimeout - } - var lock syscall.Flock_t - lock.Start = 0 - lock.Len = 0 - lock.Pid = 0 - lock.Whence = 0 - lock.Pid = 0 - if exclusive { - lock.Type = syscall.F_WRLCK - } else { - lock.Type = syscall.F_RDLCK - } - err := syscall.FcntlFlock(db.file.Fd(), syscall.F_SETLK, &lock) + // Attempt to obtain an exclusive lock. + lock := syscall.Flock_t{Type: lockType} + err := syscall.FcntlFlock(fd, syscall.F_SETLK, &lock) if err == nil { return nil } else if err != syscall.EAGAIN { return err } + // If we timed out then return an error. + if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout { + return ErrTimeout + } + // Wait for a bit and try again. - time.Sleep(50 * time.Millisecond) + time.Sleep(flockRetryTimeout) } } diff --git a/vendor/github.com/boltdb/bolt/bolt_windows.go b/vendor/go.etcd.io/bbolt/bolt_windows.go similarity index 76% rename from vendor/github.com/boltdb/bolt/bolt_windows.go rename to vendor/go.etcd.io/bbolt/bolt_windows.go index b00fb072..fca178bd 100644 --- a/vendor/github.com/boltdb/bolt/bolt_windows.go +++ b/vendor/go.etcd.io/bbolt/bolt_windows.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "fmt" @@ -16,8 +16,6 @@ var ( ) const ( - lockExt = ".lock" - // see https://msdn.microsoft.com/en-us/library/windows/desktop/aa365203(v=vs.85).aspx flagLockExclusive = 2 flagLockFailImmediately = 1 @@ -48,48 +46,47 @@ func fdatasync(db *DB) error { } // flock acquires an advisory lock on a file descriptor. -func flock(db *DB, mode os.FileMode, exclusive bool, timeout time.Duration) error { - // Create a separate lock file on windows because a process - // cannot share an exclusive lock on the same file. This is - // needed during Tx.WriteTo(). - f, err := os.OpenFile(db.path+lockExt, os.O_CREATE, mode) - if err != nil { - return err - } - db.lockfile = f - +func flock(db *DB, exclusive bool, timeout time.Duration) error { var t time.Time + if timeout != 0 { + t = time.Now() + } + var flag uint32 = flagLockFailImmediately + if exclusive { + flag |= flagLockExclusive + } for { - // If we're beyond our timeout then return an error. - // This can only occur after we've attempted a flock once. - if t.IsZero() { - t = time.Now() - } else if timeout > 0 && time.Since(t) > timeout { - return ErrTimeout - } + // Fix for https://github.com/etcd-io/bbolt/issues/121. Use byte-range + // -1..0 as the lock on the database file. + var m1 uint32 = (1 << 32) - 1 // -1 in a uint32 + err := lockFileEx(syscall.Handle(db.file.Fd()), flag, 0, 1, 0, &syscall.Overlapped{ + Offset: m1, + OffsetHigh: m1, + }) - var flag uint32 = flagLockFailImmediately - if exclusive { - flag |= flagLockExclusive - } - - err := lockFileEx(syscall.Handle(db.lockfile.Fd()), flag, 0, 1, 0, &syscall.Overlapped{}) if err == nil { return nil } else if err != errLockViolation { return err } + // If we timed oumercit then return an error. + if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout { + return ErrTimeout + } + // Wait for a bit and try again. - time.Sleep(50 * time.Millisecond) + time.Sleep(flockRetryTimeout) } } // funlock releases an advisory lock on a file descriptor. func funlock(db *DB) error { - err := unlockFileEx(syscall.Handle(db.lockfile.Fd()), 0, 1, 0, &syscall.Overlapped{}) - db.lockfile.Close() - os.Remove(db.path + lockExt) + var m1 uint32 = (1 << 32) - 1 // -1 in a uint32 + err := unlockFileEx(syscall.Handle(db.file.Fd()), 0, 1, 0, &syscall.Overlapped{ + Offset: m1, + OffsetHigh: m1, + }) return err } diff --git a/vendor/github.com/boltdb/bolt/boltsync_unix.go b/vendor/go.etcd.io/bbolt/boltsync_unix.go similarity index 91% rename from vendor/github.com/boltdb/bolt/boltsync_unix.go rename to vendor/go.etcd.io/bbolt/boltsync_unix.go index f5044252..9587afef 100644 --- a/vendor/github.com/boltdb/bolt/boltsync_unix.go +++ b/vendor/go.etcd.io/bbolt/boltsync_unix.go @@ -1,6 +1,6 @@ // +build !windows,!plan9,!linux,!openbsd -package bolt +package bbolt // fdatasync flushes written data to a file descriptor. func fdatasync(db *DB) error { diff --git a/vendor/github.com/boltdb/bolt/bucket.go b/vendor/go.etcd.io/bbolt/bucket.go similarity index 95% rename from vendor/github.com/boltdb/bolt/bucket.go rename to vendor/go.etcd.io/bbolt/bucket.go index 0c5bf274..d8750b14 100644 --- a/vendor/github.com/boltdb/bolt/bucket.go +++ b/vendor/go.etcd.io/bbolt/bucket.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "bytes" @@ -14,13 +14,6 @@ const ( MaxValueSize = (1 << 31) - 2 ) -const ( - maxUint = ^uint(0) - minUint = 0 - maxInt = int(^uint(0) >> 1) - minInt = -maxInt - 1 -) - const bucketHeaderSize = int(unsafe.Sizeof(bucket{})) const ( @@ -130,10 +123,12 @@ func (b *Bucket) Bucket(name []byte) *Bucket { func (b *Bucket) openBucket(value []byte) *Bucket { var child = newBucket(b.tx) - // If unaligned load/stores are broken on this arch and value is - // unaligned simply clone to an aligned byte array. - unaligned := brokenUnaligned && uintptr(unsafe.Pointer(&value[0]))&3 != 0 - + // Unaligned access requires a copy to be made. + const unalignedMask = unsafe.Alignof(struct { + bucket + page + }{}) - 1 + unaligned := uintptr(unsafe.Pointer(&value[0]))&unalignedMask != 0 if unaligned { value = cloneBytes(value) } @@ -213,7 +208,7 @@ func (b *Bucket) CreateBucketIfNotExists(key []byte) (*Bucket, error) { } // DeleteBucket deletes a bucket at the given key. -// Returns an error if the bucket does not exists, or if the key represents a non-bucket value. +// Returns an error if the bucket does not exist, or if the key represents a non-bucket value. func (b *Bucket) DeleteBucket(key []byte) error { if b.tx.db == nil { return ErrTxClosed @@ -235,7 +230,7 @@ func (b *Bucket) DeleteBucket(key []byte) error { // Recursively delete all child buckets. child := b.Bucket(key) err := child.ForEach(func(k, v []byte) error { - if v == nil { + if _, _, childFlags := child.Cursor().seek(k); (childFlags & bucketLeafFlag) != 0 { if err := child.DeleteBucket(k); err != nil { return fmt.Errorf("delete bucket: %s", err) } @@ -323,7 +318,12 @@ func (b *Bucket) Delete(key []byte) error { // Move cursor to correct position. c := b.Cursor() - _, _, flags := c.seek(key) + k, _, flags := c.seek(key) + + // Return nil if the key doesn't exist. + if !bytes.Equal(key, k) { + return nil + } // Return an error if there is already existing bucket value. if (flags & bucketLeafFlag) != 0 { @@ -411,7 +411,7 @@ func (b *Bucket) Stats() BucketStats { if p.count != 0 { // If page has any elements, add all element headers. - used += leafPageElementSize * int(p.count-1) + used += leafPageElementSize * uintptr(p.count-1) // Add all element key, value sizes. // The computation takes advantage of the fact that the position @@ -419,16 +419,16 @@ func (b *Bucket) Stats() BucketStats { // of all previous elements' keys and values. // It also includes the last element's header. lastElement := p.leafPageElement(p.count - 1) - used += int(lastElement.pos + lastElement.ksize + lastElement.vsize) + used += uintptr(lastElement.pos + lastElement.ksize + lastElement.vsize) } if b.root == 0 { // For inlined bucket just update the inline stats - s.InlineBucketInuse += used + s.InlineBucketInuse += int(used) } else { // For non-inlined bucket update all the leaf stats s.LeafPageN++ - s.LeafInuse += used + s.LeafInuse += int(used) s.LeafOverflowN += int(p.overflow) // Collect stats from sub-buckets. @@ -449,13 +449,13 @@ func (b *Bucket) Stats() BucketStats { // used totals the used bytes for the page // Add header and all element headers. - used := pageHeaderSize + (branchPageElementSize * int(p.count-1)) + used := pageHeaderSize + (branchPageElementSize * uintptr(p.count-1)) // Add size of all keys and values. // Again, use the fact that last element's position equals to // the total of key, value sizes of all previous elements. - used += int(lastElement.pos + lastElement.ksize) - s.BranchInuse += used + used += uintptr(lastElement.pos + lastElement.ksize) + s.BranchInuse += int(used) s.BranchOverflowN += int(p.overflow) } @@ -595,7 +595,7 @@ func (b *Bucket) inlineable() bool { // our threshold for inline bucket size. var size = pageHeaderSize for _, inode := range n.inodes { - size += leafPageElementSize + len(inode.key) + len(inode.value) + size += leafPageElementSize + uintptr(len(inode.key)) + uintptr(len(inode.value)) if inode.flags&bucketLeafFlag != 0 { return false @@ -608,8 +608,8 @@ func (b *Bucket) inlineable() bool { } // Returns the maximum total size of a bucket to make it a candidate for inlining. -func (b *Bucket) maxInlineBucketSize() int { - return b.tx.db.pageSize / 4 +func (b *Bucket) maxInlineBucketSize() uintptr { + return uintptr(b.tx.db.pageSize / 4) } // write allocates and writes a bucket to a byte slice. diff --git a/vendor/github.com/boltdb/bolt/cursor.go b/vendor/go.etcd.io/bbolt/cursor.go similarity index 98% rename from vendor/github.com/boltdb/bolt/cursor.go rename to vendor/go.etcd.io/bbolt/cursor.go index 1be9f35e..98aeb449 100644 --- a/vendor/github.com/boltdb/bolt/cursor.go +++ b/vendor/go.etcd.io/bbolt/cursor.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "bytes" @@ -157,12 +157,6 @@ func (c *Cursor) seek(seek []byte) (key []byte, value []byte, flags uint32) { // Start from root page/node and traverse to correct page. c.stack = c.stack[:0] c.search(seek, c.bucket.root) - ref := &c.stack[len(c.stack)-1] - - // If the cursor is pointing to the end of page/node then return nil. - if ref.index >= ref.count() { - return nil, nil, 0 - } // If this is a bucket then return a nil value. return c.keyValue() @@ -339,6 +333,8 @@ func (c *Cursor) nsearch(key []byte) { // keyValue returns the key and value of the current leaf element. func (c *Cursor) keyValue() ([]byte, []byte, uint32) { ref := &c.stack[len(c.stack)-1] + + // If the cursor is pointing to the end of page/node then return nil. if ref.count() == 0 || ref.index >= ref.count() { return nil, nil, 0 } @@ -370,7 +366,7 @@ func (c *Cursor) node() *node { } for _, ref := range c.stack[:len(c.stack)-1] { _assert(!n.isLeaf, "expected branch node") - n = n.childAt(int(ref.index)) + n = n.childAt(ref.index) } _assert(n.isLeaf, "expected leaf node") return n diff --git a/vendor/github.com/boltdb/bolt/db.go b/vendor/go.etcd.io/bbolt/db.go similarity index 80% rename from vendor/github.com/boltdb/bolt/db.go rename to vendor/go.etcd.io/bbolt/db.go index f352ff14..80b0095c 100644 --- a/vendor/github.com/boltdb/bolt/db.go +++ b/vendor/go.etcd.io/bbolt/db.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "errors" @@ -7,8 +7,7 @@ import ( "log" "os" "runtime" - "runtime/debug" - "strings" + "sort" "sync" "time" "unsafe" @@ -23,6 +22,8 @@ const version = 2 // Represents a marker value to indicate that a file is a Bolt DB. const magic uint32 = 0xED0CDAED +const pgidNoFreelist pgid = 0xffffffffffffffff + // IgnoreNoSync specifies whether the NoSync field of a DB is ignored when // syncing changes to a file. This is required as some operating systems, // such as OpenBSD, do not have a unified buffer cache (UBC) and writes @@ -39,6 +40,19 @@ const ( // default page size for db is set to the OS page size. var defaultPageSize = os.Getpagesize() +// The time elapsed between consecutive file locking attempts. +const flockRetryTimeout = 50 * time.Millisecond + +// FreelistType is the type of the freelist backend +type FreelistType string + +const ( + // FreelistArrayType indicates backend freelist type is array + FreelistArrayType = FreelistType("array") + // FreelistMapType indicates backend freelist type is hashmap + FreelistMapType = FreelistType("hashmap") +) + // DB represents a collection of buckets persisted to a file on disk. // All data access is performed through transactions which can be obtained through the DB. // All the functions on DB will return a ErrDatabaseNotOpen if accessed before Open() is called. @@ -61,6 +75,18 @@ type DB struct { // THIS IS UNSAFE. PLEASE USE WITH CAUTION. NoSync bool + // When true, skips syncing freelist to disk. This improves the database + // write performance under normal operation, but requires a full database + // re-sync during recovery. + NoFreelistSync bool + + // FreelistType sets the backend freelist type. There are two options. Array which is simple but endures + // dramatic performance degradation if database is large and framentation in freelist is common. + // The alternative one is using hashmap, it is faster in almost all circumstances + // but it doesn't guarantee that it offers the smallest page id available. In normal case it is safe. + // The default type is array + FreelistType FreelistType + // When true, skips the truncate call when growing the database. // Setting this to true is only safe on non-ext3/ext4 systems. // Skipping truncation avoids preallocation of hard drive space and @@ -95,9 +121,9 @@ type DB struct { AllocSize int path string + openFile func(string, int, os.FileMode) (*os.File, error) file *os.File - lockfile *os.File // windows only - dataref []byte // mmap'ed readonly, write throws SEGV + dataref []byte // mmap'ed readonly, write throws SEGV data *[maxMapSize]byte datasz int filesz int // current on disk file size @@ -107,9 +133,11 @@ type DB struct { opened bool rwtx *Tx txs []*Tx - freelist *freelist stats Stats + freelist *freelist + freelistLoad sync.Once + pagePool sync.Pool batchMu sync.Mutex @@ -148,14 +176,18 @@ func (db *DB) String() string { // If the file does not exist then it will be created automatically. // Passing in nil options will cause Bolt to open the database with the default options. func Open(path string, mode os.FileMode, options *Options) (*DB, error) { - var db = &DB{opened: true} - + db := &DB{ + opened: true, + } // Set default options if no options are provided. if options == nil { options = DefaultOptions } + db.NoSync = options.NoSync db.NoGrowSync = options.NoGrowSync db.MmapFlags = options.MmapFlags + db.NoFreelistSync = options.NoFreelistSync + db.FreelistType = options.FreelistType // Set default values for later DB operations. db.MaxBatchSize = DefaultMaxBatchSize @@ -168,13 +200,18 @@ func Open(path string, mode os.FileMode, options *Options) (*DB, error) { db.readOnly = true } + db.openFile = options.OpenFile + if db.openFile == nil { + db.openFile = os.OpenFile + } + // Open data file and separate sync handler for metadata writes. - db.path = path var err error - if db.file, err = os.OpenFile(db.path, flag|os.O_CREATE, mode); err != nil { + if db.file, err = db.openFile(path, flag|os.O_CREATE, mode); err != nil { _ = db.close() return nil, err } + db.path = db.file.Name() // Lock file so that other processes using Bolt in read-write mode cannot // use the database at the same time. This would cause corruption since @@ -183,7 +220,7 @@ func Open(path string, mode os.FileMode, options *Options) (*DB, error) { // if !options.ReadOnly. // The database file is locked using the shared lock (more than one process may // hold a lock at the same time) otherwise (options.ReadOnly is set). - if err := flock(db, mode, !db.readOnly, options.Timeout); err != nil { + if err := flock(db, !db.readOnly, options.Timeout); err != nil { _ = db.close() return nil, err } @@ -191,31 +228,41 @@ func Open(path string, mode os.FileMode, options *Options) (*DB, error) { // Default values for test hooks db.ops.writeAt = db.file.WriteAt + if db.pageSize = options.PageSize; db.pageSize == 0 { + // Set the default page size to the OS page size. + db.pageSize = defaultPageSize + } + // Initialize the database if it doesn't exist. if info, err := db.file.Stat(); err != nil { + _ = db.close() return nil, err } else if info.Size() == 0 { // Initialize new files with meta pages. if err := db.init(); err != nil { + // clean up file descriptor on initialization fail + _ = db.close() return nil, err } } else { // Read the first meta page to determine the page size. var buf [0x1000]byte - if _, err := db.file.ReadAt(buf[:], 0); err == nil { - m := db.pageInBuffer(buf[:], 0).meta() - if err := m.validate(); err != nil { - // If we can't read the page size, we can assume it's the same - // as the OS -- since that's how the page size was chosen in the - // first place. - // - // If the first page is invalid and this OS uses a different - // page size than what the database was created with then we - // are out of luck and cannot access the database. - db.pageSize = os.Getpagesize() - } else { + // If we can't read the page size, but can read a page, assume + // it's the same as the OS or one given -- since that's how the + // page size was chosen in the first place. + // + // If the first page is invalid and this OS uses a different + // page size than what the database was created with then we + // are out of luck and cannot access the database. + // + // TODO: scan for next page + if bw, err := db.file.ReadAt(buf[:], 0); err == nil && bw == len(buf) { + if m := db.pageInBuffer(buf[:], 0).meta(); m.validate() == nil { db.pageSize = int(m.pageSize) } + } else { + _ = db.close() + return nil, ErrInvalid } } @@ -232,14 +279,50 @@ func Open(path string, mode os.FileMode, options *Options) (*DB, error) { return nil, err } - // Read in the freelist. - db.freelist = newFreelist() - db.freelist.read(db.page(db.meta().freelist)) + if db.readOnly { + return db, nil + } + + db.loadFreelist() + + // Flush freelist when transitioning from no sync to sync so + // NoFreelistSync unaware boltdb can open the db later. + if !db.NoFreelistSync && !db.hasSyncedFreelist() { + tx, err := db.Begin(true) + if tx != nil { + err = tx.Commit() + } + if err != nil { + _ = db.close() + return nil, err + } + } // Mark the database as opened and return. return db, nil } +// loadFreelist reads the freelist if it is synced, or reconstructs it +// by scanning the DB if it is not synced. It assumes there are no +// concurrent accesses being made to the freelist. +func (db *DB) loadFreelist() { + db.freelistLoad.Do(func() { + db.freelist = newFreelist(db.FreelistType) + if !db.hasSyncedFreelist() { + // Reconstruct free list by scanning the DB. + db.freelist.readIDs(db.freepages()) + } else { + // Read free list from freelist page. + db.freelist.read(db.page(db.meta().freelist)) + } + db.stats.FreePageN = db.freelist.free_count() + }) +} + +func (db *DB) hasSyncedFreelist() bool { + return db.meta().freelist != pgidNoFreelist +} + // mmap opens the underlying memory-mapped file and initializes the meta references. // minsz is the minimum size that the new mmap can be. func (db *DB) mmap(minsz int) error { @@ -341,9 +424,6 @@ func (db *DB) mmapSize(size int) (int, error) { // init creates a new database file and initializes its meta pages. func (db *DB) init() error { - // Set the page size to the OS page size. - db.pageSize = os.Getpagesize() - // Create two meta pages on a buffer. buf := make([]byte, db.pageSize*4) for i := 0; i < 2; i++ { @@ -387,7 +467,8 @@ func (db *DB) init() error { } // Close releases all database resources. -// All transactions must be closed before closing the database. +// It will block waiting for any open transactions to finish +// before closing the database and returning. func (db *DB) Close() error { db.rwlock.Lock() defer db.rwlock.Unlock() @@ -395,8 +476,8 @@ func (db *DB) Close() error { db.metalock.Lock() defer db.metalock.Unlock() - db.mmaplock.RLock() - defer db.mmaplock.RUnlock() + db.mmaplock.Lock() + defer db.mmaplock.Unlock() return db.close() } @@ -526,21 +607,36 @@ func (db *DB) beginRWTx() (*Tx, error) { t := &Tx{writable: true} t.init(db) db.rwtx = t + db.freePages() + return t, nil +} - // Free any pages associated with closed read-only transactions. - var minid txid = 0xFFFFFFFFFFFFFFFF - for _, t := range db.txs { - if t.meta.txid < minid { - minid = t.meta.txid - } +// freePages releases any pages associated with closed read-only transactions. +func (db *DB) freePages() { + // Free all pending pages prior to earliest open transaction. + sort.Sort(txsById(db.txs)) + minid := txid(0xFFFFFFFFFFFFFFFF) + if len(db.txs) > 0 { + minid = db.txs[0].meta.txid } if minid > 0 { db.freelist.release(minid - 1) } - - return t, nil + // Release unused txid extents. + for _, t := range db.txs { + db.freelist.releaseRange(minid, t.meta.txid-1) + minid = t.meta.txid + 1 + } + db.freelist.releaseRange(minid, txid(0xFFFFFFFFFFFFFFFF)) + // Any page both allocated and freed in an extent is safe to release. } +type txsById []*Tx + +func (t txsById) Len() int { return len(t) } +func (t txsById) Swap(i, j int) { t[i], t[j] = t[j], t[i] } +func (t txsById) Less(i, j int) bool { return t[i].meta.txid < t[j].meta.txid } + // removeTx removes a transaction from the database. func (db *DB) removeTx(tx *Tx) { // Release the read lock on the mmap. @@ -633,11 +729,7 @@ func (db *DB) View(fn func(*Tx) error) error { return err } - if err := t.Rollback(); err != nil { - return err - } - - return nil + return t.Rollback() } // Batch calls fn as part of a batch. It behaves similar to Update, @@ -737,9 +829,7 @@ retry: // pass success, or bolt internal errors, to all callers for _, c := range b.calls { - if c.err != nil { - c.err <- err - } + c.err <- err } break retry } @@ -826,7 +916,7 @@ func (db *DB) meta() *meta { } // allocate returns a contiguous block of memory starting at a given page. -func (db *DB) allocate(count int) (*page, error) { +func (db *DB) allocate(txid txid, count int) (*page, error) { // Allocate a temporary buffer for the page. var buf []byte if count == 1 { @@ -838,7 +928,7 @@ func (db *DB) allocate(count int) (*page, error) { p.overflow = uint32(count - 1) // Use pages from the freelist if they are available. - if p.id = db.freelist.allocate(count); p.id != 0 { + if p.id = db.freelist.allocate(txid, count); p.id != 0 { return p, nil } @@ -893,6 +983,38 @@ func (db *DB) IsReadOnly() bool { return db.readOnly } +func (db *DB) freepages() []pgid { + tx, err := db.beginTx() + defer func() { + err = tx.Rollback() + if err != nil { + panic("freepages: failed to rollback tx") + } + }() + if err != nil { + panic("freepages: failed to open read only tx") + } + + reachable := make(map[pgid]*page) + nofreed := make(map[pgid]bool) + ech := make(chan error) + go func() { + for e := range ech { + panic(fmt.Sprintf("freepages: failed to get all reachable pages (%v)", e)) + } + }() + tx.checkBucket(&tx.root, reachable, nofreed, ech) + close(ech) + + var fids []pgid + for i := pgid(2); i < db.meta().pgid; i++ { + if _, ok := reachable[i]; !ok { + fids = append(fids, i) + } + } + return fids +} + // Options represents the options that can be set when opening a database. type Options struct { // Timeout is the amount of time to wait to obtain a file lock. @@ -903,6 +1025,17 @@ type Options struct { // Sets the DB.NoGrowSync flag before memory mapping the file. NoGrowSync bool + // Do not sync freelist to disk. This improves the database write performance + // under normal operation, but requires a full database re-sync during recovery. + NoFreelistSync bool + + // FreelistType sets the backend freelist type. There are two options. Array which is simple but endures + // dramatic performance degradation if database is large and framentation in freelist is common. + // The alternative one is using hashmap, it is faster in almost all circumstances + // but it doesn't guarantee that it offers the smallest page id available. In normal case it is safe. + // The default type is array + FreelistType FreelistType + // Open database in read-only mode. Uses flock(..., LOCK_SH |LOCK_NB) to // grab a shared lock (UNIX). ReadOnly bool @@ -919,13 +1052,26 @@ type Options struct { // If initialMmapSize is smaller than the previous database size, // it takes no effect. InitialMmapSize int + + // PageSize overrides the default OS page size. + PageSize int + + // NoSync sets the initial value of DB.NoSync. Normally this can just be + // set directly on the DB itself when returned from Open(), but this option + // is useful in APIs which expose Options but not the underlying DB. + NoSync bool + + // OpenFile is used to open files. It defaults to os.OpenFile. This option + // is useful for writing hermetic tests. + OpenFile func(string, int, os.FileMode) (*os.File, error) } // DefaultOptions represent the options used if nil options are passed into Open(). // No timeout is used which will cause Bolt to wait indefinitely for a lock. var DefaultOptions = &Options{ - Timeout: 0, - NoGrowSync: false, + Timeout: 0, + NoGrowSync: false, + FreelistType: FreelistArrayType, } // Stats represents statistics about the database. @@ -960,10 +1106,6 @@ func (s *Stats) Sub(other *Stats) Stats { return diff } -func (s *Stats) add(other *Stats) { - s.TxStats.add(&other.TxStats) -} - type Info struct { Data uintptr PageSize int @@ -1002,7 +1144,8 @@ func (m *meta) copy(dest *meta) { func (m *meta) write(p *page) { if m.root.root >= m.pgid { panic(fmt.Sprintf("root bucket pgid (%d) above high water mark (%d)", m.root.root, m.pgid)) - } else if m.freelist >= m.pgid { + } else if m.freelist >= m.pgid && m.freelist != pgidNoFreelist { + // TODO: reject pgidNoFreeList if !NoFreelistSync panic(fmt.Sprintf("freelist pgid (%d) above high water mark (%d)", m.freelist, m.pgid)) } @@ -1029,11 +1172,3 @@ func _assert(condition bool, msg string, v ...interface{}) { panic(fmt.Sprintf("assertion failed: "+msg, v...)) } } - -func warn(v ...interface{}) { fmt.Fprintln(os.Stderr, v...) } -func warnf(msg string, v ...interface{}) { fmt.Fprintf(os.Stderr, msg+"\n", v...) } - -func printstack() { - stack := strings.Join(strings.Split(string(debug.Stack()), "\n")[2:], "\n") - fmt.Fprintln(os.Stderr, stack) -} diff --git a/vendor/github.com/boltdb/bolt/doc.go b/vendor/go.etcd.io/bbolt/doc.go similarity index 94% rename from vendor/github.com/boltdb/bolt/doc.go rename to vendor/go.etcd.io/bbolt/doc.go index cc937845..95f25f01 100644 --- a/vendor/github.com/boltdb/bolt/doc.go +++ b/vendor/go.etcd.io/bbolt/doc.go @@ -1,5 +1,5 @@ /* -Package bolt implements a low-level key/value store in pure Go. It supports +package bbolt implements a low-level key/value store in pure Go. It supports fully serializable transactions, ACID semantics, and lock-free MVCC with multiple readers and a single writer. Bolt can be used for projects that want a simple data store without the need to add large dependencies such as @@ -41,4 +41,4 @@ point to different data or can point to invalid memory which will cause a panic. */ -package bolt +package bbolt diff --git a/vendor/github.com/boltdb/bolt/errors.go b/vendor/go.etcd.io/bbolt/errors.go similarity index 99% rename from vendor/github.com/boltdb/bolt/errors.go rename to vendor/go.etcd.io/bbolt/errors.go index a3620a3e..48758ca5 100644 --- a/vendor/github.com/boltdb/bolt/errors.go +++ b/vendor/go.etcd.io/bbolt/errors.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import "errors" diff --git a/vendor/go.etcd.io/bbolt/freelist.go b/vendor/go.etcd.io/bbolt/freelist.go new file mode 100644 index 00000000..697a4696 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/freelist.go @@ -0,0 +1,404 @@ +package bbolt + +import ( + "fmt" + "sort" + "unsafe" +) + +// txPending holds a list of pgids and corresponding allocation txns +// that are pending to be freed. +type txPending struct { + ids []pgid + alloctx []txid // txids allocating the ids + lastReleaseBegin txid // beginning txid of last matching releaseRange +} + +// pidSet holds the set of starting pgids which have the same span size +type pidSet map[pgid]struct{} + +// freelist represents a list of all pages that are available for allocation. +// It also tracks pages that have been freed but are still in use by open transactions. +type freelist struct { + freelistType FreelistType // freelist type + ids []pgid // all free and available free page ids. + allocs map[pgid]txid // mapping of txid that allocated a pgid. + pending map[txid]*txPending // mapping of soon-to-be free page ids by tx. + cache map[pgid]bool // fast lookup of all free and pending page ids. + freemaps map[uint64]pidSet // key is the size of continuous pages(span), value is a set which contains the starting pgids of same size + forwardMap map[pgid]uint64 // key is start pgid, value is its span size + backwardMap map[pgid]uint64 // key is end pgid, value is its span size + allocate func(txid txid, n int) pgid // the freelist allocate func + free_count func() int // the function which gives you free page number + mergeSpans func(ids pgids) // the mergeSpan func + getFreePageIDs func() []pgid // get free pgids func + readIDs func(pgids []pgid) // readIDs func reads list of pages and init the freelist +} + +// newFreelist returns an empty, initialized freelist. +func newFreelist(freelistType FreelistType) *freelist { + f := &freelist{ + freelistType: freelistType, + allocs: make(map[pgid]txid), + pending: make(map[txid]*txPending), + cache: make(map[pgid]bool), + freemaps: make(map[uint64]pidSet), + forwardMap: make(map[pgid]uint64), + backwardMap: make(map[pgid]uint64), + } + + if freelistType == FreelistMapType { + f.allocate = f.hashmapAllocate + f.free_count = f.hashmapFreeCount + f.mergeSpans = f.hashmapMergeSpans + f.getFreePageIDs = f.hashmapGetFreePageIDs + f.readIDs = f.hashmapReadIDs + } else { + f.allocate = f.arrayAllocate + f.free_count = f.arrayFreeCount + f.mergeSpans = f.arrayMergeSpans + f.getFreePageIDs = f.arrayGetFreePageIDs + f.readIDs = f.arrayReadIDs + } + + return f +} + +// size returns the size of the page after serialization. +func (f *freelist) size() int { + n := f.count() + if n >= 0xFFFF { + // The first element will be used to store the count. See freelist.write. + n++ + } + return int(pageHeaderSize) + (int(unsafe.Sizeof(pgid(0))) * n) +} + +// count returns count of pages on the freelist +func (f *freelist) count() int { + return f.free_count() + f.pending_count() +} + +// arrayFreeCount returns count of free pages(array version) +func (f *freelist) arrayFreeCount() int { + return len(f.ids) +} + +// pending_count returns count of pending pages +func (f *freelist) pending_count() int { + var count int + for _, txp := range f.pending { + count += len(txp.ids) + } + return count +} + +// copyall copies a list of all free ids and all pending ids in one sorted list. +// f.count returns the minimum length required for dst. +func (f *freelist) copyall(dst []pgid) { + m := make(pgids, 0, f.pending_count()) + for _, txp := range f.pending { + m = append(m, txp.ids...) + } + sort.Sort(m) + mergepgids(dst, f.getFreePageIDs(), m) +} + +// arrayAllocate returns the starting page id of a contiguous list of pages of a given size. +// If a contiguous block cannot be found then 0 is returned. +func (f *freelist) arrayAllocate(txid txid, n int) pgid { + if len(f.ids) == 0 { + return 0 + } + + var initial, previd pgid + for i, id := range f.ids { + if id <= 1 { + panic(fmt.Sprintf("invalid page allocation: %d", id)) + } + + // Reset initial page if this is not contiguous. + if previd == 0 || id-previd != 1 { + initial = id + } + + // If we found a contiguous block then remove it and return it. + if (id-initial)+1 == pgid(n) { + // If we're allocating off the beginning then take the fast path + // and just adjust the existing slice. This will use extra memory + // temporarily but the append() in free() will realloc the slice + // as is necessary. + if (i + 1) == n { + f.ids = f.ids[i+1:] + } else { + copy(f.ids[i-n+1:], f.ids[i+1:]) + f.ids = f.ids[:len(f.ids)-n] + } + + // Remove from the free cache. + for i := pgid(0); i < pgid(n); i++ { + delete(f.cache, initial+i) + } + f.allocs[initial] = txid + return initial + } + + previd = id + } + return 0 +} + +// free releases a page and its overflow for a given transaction id. +// If the page is already free then a panic will occur. +func (f *freelist) free(txid txid, p *page) { + if p.id <= 1 { + panic(fmt.Sprintf("cannot free page 0 or 1: %d", p.id)) + } + + // Free page and all its overflow pages. + txp := f.pending[txid] + if txp == nil { + txp = &txPending{} + f.pending[txid] = txp + } + allocTxid, ok := f.allocs[p.id] + if ok { + delete(f.allocs, p.id) + } else if (p.flags & freelistPageFlag) != 0 { + // Freelist is always allocated by prior tx. + allocTxid = txid - 1 + } + + for id := p.id; id <= p.id+pgid(p.overflow); id++ { + // Verify that page is not already free. + if f.cache[id] { + panic(fmt.Sprintf("page %d already freed", id)) + } + // Add to the freelist and cache. + txp.ids = append(txp.ids, id) + txp.alloctx = append(txp.alloctx, allocTxid) + f.cache[id] = true + } +} + +// release moves all page ids for a transaction id (or older) to the freelist. +func (f *freelist) release(txid txid) { + m := make(pgids, 0) + for tid, txp := range f.pending { + if tid <= txid { + // Move transaction's pending pages to the available freelist. + // Don't remove from the cache since the page is still free. + m = append(m, txp.ids...) + delete(f.pending, tid) + } + } + f.mergeSpans(m) +} + +// releaseRange moves pending pages allocated within an extent [begin,end] to the free list. +func (f *freelist) releaseRange(begin, end txid) { + if begin > end { + return + } + var m pgids + for tid, txp := range f.pending { + if tid < begin || tid > end { + continue + } + // Don't recompute freed pages if ranges haven't updated. + if txp.lastReleaseBegin == begin { + continue + } + for i := 0; i < len(txp.ids); i++ { + if atx := txp.alloctx[i]; atx < begin || atx > end { + continue + } + m = append(m, txp.ids[i]) + txp.ids[i] = txp.ids[len(txp.ids)-1] + txp.ids = txp.ids[:len(txp.ids)-1] + txp.alloctx[i] = txp.alloctx[len(txp.alloctx)-1] + txp.alloctx = txp.alloctx[:len(txp.alloctx)-1] + i-- + } + txp.lastReleaseBegin = begin + if len(txp.ids) == 0 { + delete(f.pending, tid) + } + } + f.mergeSpans(m) +} + +// rollback removes the pages from a given pending tx. +func (f *freelist) rollback(txid txid) { + // Remove page ids from cache. + txp := f.pending[txid] + if txp == nil { + return + } + var m pgids + for i, pgid := range txp.ids { + delete(f.cache, pgid) + tx := txp.alloctx[i] + if tx == 0 { + continue + } + if tx != txid { + // Pending free aborted; restore page back to alloc list. + f.allocs[pgid] = tx + } else { + // Freed page was allocated by this txn; OK to throw away. + m = append(m, pgid) + } + } + // Remove pages from pending list and mark as free if allocated by txid. + delete(f.pending, txid) + f.mergeSpans(m) +} + +// freed returns whether a given page is in the free list. +func (f *freelist) freed(pgid pgid) bool { + return f.cache[pgid] +} + +// read initializes the freelist from a freelist page. +func (f *freelist) read(p *page) { + if (p.flags & freelistPageFlag) == 0 { + panic(fmt.Sprintf("invalid freelist page: %d, page type is %s", p.id, p.typ())) + } + // If the page.count is at the max uint16 value (64k) then it's considered + // an overflow and the size of the freelist is stored as the first element. + var idx, count = 0, int(p.count) + if count == 0xFFFF { + idx = 1 + c := *(*pgid)(unsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p))) + count = int(c) + if count < 0 { + panic(fmt.Sprintf("leading element count %d overflows int", c)) + } + } + + // Copy the list of page ids from the freelist. + if count == 0 { + f.ids = nil + } else { + var ids []pgid + data := unsafeIndex(unsafe.Pointer(p), unsafe.Sizeof(*p), unsafe.Sizeof(ids[0]), idx) + unsafeSlice(unsafe.Pointer(&ids), data, count) + + // copy the ids, so we don't modify on the freelist page directly + idsCopy := make([]pgid, count) + copy(idsCopy, ids) + // Make sure they're sorted. + sort.Sort(pgids(idsCopy)) + + f.readIDs(idsCopy) + } +} + +// arrayReadIDs initializes the freelist from a given list of ids. +func (f *freelist) arrayReadIDs(ids []pgid) { + f.ids = ids + f.reindex() +} + +func (f *freelist) arrayGetFreePageIDs() []pgid { + return f.ids +} + +// write writes the page ids onto a freelist page. All free and pending ids are +// saved to disk since in the event of a program crash, all pending ids will +// become free. +func (f *freelist) write(p *page) error { + // Combine the old free pgids and pgids waiting on an open transaction. + + // Update the header flag. + p.flags |= freelistPageFlag + + // The page.count can only hold up to 64k elements so if we overflow that + // number then we handle it by putting the size in the first element. + l := f.count() + if l == 0 { + p.count = uint16(l) + } else if l < 0xFFFF { + p.count = uint16(l) + var ids []pgid + data := unsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p)) + unsafeSlice(unsafe.Pointer(&ids), data, l) + f.copyall(ids) + } else { + p.count = 0xFFFF + var ids []pgid + data := unsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p)) + unsafeSlice(unsafe.Pointer(&ids), data, l+1) + ids[0] = pgid(l) + f.copyall(ids[1:]) + } + + return nil +} + +// reload reads the freelist from a page and filters out pending items. +func (f *freelist) reload(p *page) { + f.read(p) + + // Build a cache of only pending pages. + pcache := make(map[pgid]bool) + for _, txp := range f.pending { + for _, pendingID := range txp.ids { + pcache[pendingID] = true + } + } + + // Check each page in the freelist and build a new available freelist + // with any pages not in the pending lists. + var a []pgid + for _, id := range f.getFreePageIDs() { + if !pcache[id] { + a = append(a, id) + } + } + + f.readIDs(a) +} + +// noSyncReload reads the freelist from pgids and filters out pending items. +func (f *freelist) noSyncReload(pgids []pgid) { + // Build a cache of only pending pages. + pcache := make(map[pgid]bool) + for _, txp := range f.pending { + for _, pendingID := range txp.ids { + pcache[pendingID] = true + } + } + + // Check each page in the freelist and build a new available freelist + // with any pages not in the pending lists. + var a []pgid + for _, id := range pgids { + if !pcache[id] { + a = append(a, id) + } + } + + f.readIDs(a) +} + +// reindex rebuilds the free cache based on available and pending free lists. +func (f *freelist) reindex() { + ids := f.getFreePageIDs() + f.cache = make(map[pgid]bool, len(ids)) + for _, id := range ids { + f.cache[id] = true + } + for _, txp := range f.pending { + for _, pendingID := range txp.ids { + f.cache[pendingID] = true + } + } +} + +// arrayMergeSpans try to merge list of pages(represented by pgids) with existing spans but using array +func (f *freelist) arrayMergeSpans(ids pgids) { + sort.Sort(ids) + f.ids = pgids(f.ids).merge(ids) +} diff --git a/vendor/go.etcd.io/bbolt/freelist_hmap.go b/vendor/go.etcd.io/bbolt/freelist_hmap.go new file mode 100644 index 00000000..02ef2be0 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/freelist_hmap.go @@ -0,0 +1,178 @@ +package bbolt + +import "sort" + +// hashmapFreeCount returns count of free pages(hashmap version) +func (f *freelist) hashmapFreeCount() int { + // use the forwardmap to get the total count + count := 0 + for _, size := range f.forwardMap { + count += int(size) + } + return count +} + +// hashmapAllocate serves the same purpose as arrayAllocate, but use hashmap as backend +func (f *freelist) hashmapAllocate(txid txid, n int) pgid { + if n == 0 { + return 0 + } + + // if we have a exact size match just return short path + if bm, ok := f.freemaps[uint64(n)]; ok { + for pid := range bm { + // remove the span + f.delSpan(pid, uint64(n)) + + f.allocs[pid] = txid + + for i := pgid(0); i < pgid(n); i++ { + delete(f.cache, pid+i) + } + return pid + } + } + + // lookup the map to find larger span + for size, bm := range f.freemaps { + if size < uint64(n) { + continue + } + + for pid := range bm { + // remove the initial + f.delSpan(pid, uint64(size)) + + f.allocs[pid] = txid + + remain := size - uint64(n) + + // add remain span + f.addSpan(pid+pgid(n), remain) + + for i := pgid(0); i < pgid(n); i++ { + delete(f.cache, pid+pgid(i)) + } + return pid + } + } + + return 0 +} + +// hashmapReadIDs reads pgids as input an initial the freelist(hashmap version) +func (f *freelist) hashmapReadIDs(pgids []pgid) { + f.init(pgids) + + // Rebuild the page cache. + f.reindex() +} + +// hashmapGetFreePageIDs returns the sorted free page ids +func (f *freelist) hashmapGetFreePageIDs() []pgid { + count := f.free_count() + if count == 0 { + return nil + } + + m := make([]pgid, 0, count) + for start, size := range f.forwardMap { + for i := 0; i < int(size); i++ { + m = append(m, start+pgid(i)) + } + } + sort.Sort(pgids(m)) + + return m +} + +// hashmapMergeSpans try to merge list of pages(represented by pgids) with existing spans +func (f *freelist) hashmapMergeSpans(ids pgids) { + for _, id := range ids { + // try to see if we can merge and update + f.mergeWithExistingSpan(id) + } +} + +// mergeWithExistingSpan merges pid to the existing free spans, try to merge it backward and forward +func (f *freelist) mergeWithExistingSpan(pid pgid) { + prev := pid - 1 + next := pid + 1 + + preSize, mergeWithPrev := f.backwardMap[prev] + nextSize, mergeWithNext := f.forwardMap[next] + newStart := pid + newSize := uint64(1) + + if mergeWithPrev { + //merge with previous span + start := prev + 1 - pgid(preSize) + f.delSpan(start, preSize) + + newStart -= pgid(preSize) + newSize += preSize + } + + if mergeWithNext { + // merge with next span + f.delSpan(next, nextSize) + newSize += nextSize + } + + f.addSpan(newStart, newSize) +} + +func (f *freelist) addSpan(start pgid, size uint64) { + f.backwardMap[start-1+pgid(size)] = size + f.forwardMap[start] = size + if _, ok := f.freemaps[size]; !ok { + f.freemaps[size] = make(map[pgid]struct{}) + } + + f.freemaps[size][start] = struct{}{} +} + +func (f *freelist) delSpan(start pgid, size uint64) { + delete(f.forwardMap, start) + delete(f.backwardMap, start+pgid(size-1)) + delete(f.freemaps[size], start) + if len(f.freemaps[size]) == 0 { + delete(f.freemaps, size) + } +} + +// initial from pgids using when use hashmap version +// pgids must be sorted +func (f *freelist) init(pgids []pgid) { + if len(pgids) == 0 { + return + } + + size := uint64(1) + start := pgids[0] + + if !sort.SliceIsSorted([]pgid(pgids), func(i, j int) bool { return pgids[i] < pgids[j] }) { + panic("pgids not sorted") + } + + f.freemaps = make(map[uint64]pidSet) + f.forwardMap = make(map[pgid]uint64) + f.backwardMap = make(map[pgid]uint64) + + for i := 1; i < len(pgids); i++ { + // continuous page + if pgids[i] == pgids[i-1]+1 { + size++ + } else { + f.addSpan(start, size) + + size = 1 + start = pgids[i] + } + } + + // init the tail + if size != 0 && start != 0 { + f.addSpan(start, size) + } +} diff --git a/vendor/go.etcd.io/bbolt/go.mod b/vendor/go.etcd.io/bbolt/go.mod new file mode 100644 index 00000000..c2366dae --- /dev/null +++ b/vendor/go.etcd.io/bbolt/go.mod @@ -0,0 +1,5 @@ +module go.etcd.io/bbolt + +go 1.12 + +require golang.org/x/sys v0.0.0-20200202164722-d101bd2416d5 diff --git a/vendor/go.etcd.io/bbolt/go.sum b/vendor/go.etcd.io/bbolt/go.sum new file mode 100644 index 00000000..4ad15a48 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/go.sum @@ -0,0 +1,2 @@ +golang.org/x/sys v0.0.0-20200202164722-d101bd2416d5 h1:LfCXLvNmTYH9kEmVgqbnsWfruoXZIrh4YBgqVHtDvw0= +golang.org/x/sys v0.0.0-20200202164722-d101bd2416d5/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= diff --git a/vendor/github.com/boltdb/bolt/node.go b/vendor/go.etcd.io/bbolt/node.go similarity index 91% rename from vendor/github.com/boltdb/bolt/node.go rename to vendor/go.etcd.io/bbolt/node.go index 159318b2..73988b5c 100644 --- a/vendor/github.com/boltdb/bolt/node.go +++ b/vendor/go.etcd.io/bbolt/node.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "bytes" @@ -41,19 +41,19 @@ func (n *node) size() int { sz, elsz := pageHeaderSize, n.pageElementSize() for i := 0; i < len(n.inodes); i++ { item := &n.inodes[i] - sz += elsz + len(item.key) + len(item.value) + sz += elsz + uintptr(len(item.key)) + uintptr(len(item.value)) } - return sz + return int(sz) } // sizeLessThan returns true if the node is less than a given size. // This is an optimization to avoid calculating a large node when we only need // to know if it fits inside a certain page size. -func (n *node) sizeLessThan(v int) bool { +func (n *node) sizeLessThan(v uintptr) bool { sz, elsz := pageHeaderSize, n.pageElementSize() for i := 0; i < len(n.inodes); i++ { item := &n.inodes[i] - sz += elsz + len(item.key) + len(item.value) + sz += elsz + uintptr(len(item.key)) + uintptr(len(item.value)) if sz >= v { return false } @@ -62,7 +62,7 @@ func (n *node) sizeLessThan(v int) bool { } // pageElementSize returns the size of each page element based on the type of node. -func (n *node) pageElementSize() int { +func (n *node) pageElementSize() uintptr { if n.isLeaf { return leafPageElementSize } @@ -207,10 +207,17 @@ func (n *node) write(p *page) { } // Loop over each item and write it to the page. - b := (*[maxAllocSize]byte)(unsafe.Pointer(&p.ptr))[n.pageElementSize()*len(n.inodes):] + // off tracks the offset into p of the start of the next data. + off := unsafe.Sizeof(*p) + n.pageElementSize()*uintptr(len(n.inodes)) for i, item := range n.inodes { _assert(len(item.key) > 0, "write: zero-length inode key") + // Create a slice to write into of needed size and advance + // byte pointer for next iteration. + sz := len(item.key) + len(item.value) + b := unsafeByteSlice(unsafe.Pointer(p), off, 0, sz) + off += uintptr(sz) + // Write the page element. if n.isLeaf { elem := p.leafPageElement(uint16(i)) @@ -226,20 +233,9 @@ func (n *node) write(p *page) { _assert(elem.pgid != p.id, "write: circular dependency occurred") } - // If the length of key+value is larger than the max allocation size - // then we need to reallocate the byte array pointer. - // - // See: https://github.com/boltdb/bolt/pull/335 - klen, vlen := len(item.key), len(item.value) - if len(b) < klen+vlen { - b = (*[maxAllocSize]byte)(unsafe.Pointer(&b[0]))[:] - } - // Write data for the element to the end of the page. - copy(b[0:], item.key) - b = b[klen:] - copy(b[0:], item.value) - b = b[vlen:] + l := copy(b, item.key) + copy(b[l:], item.value) } // DEBUG ONLY: n.dump() @@ -247,7 +243,7 @@ func (n *node) write(p *page) { // split breaks up a node into multiple smaller nodes, if appropriate. // This should only be called from the spill() function. -func (n *node) split(pageSize int) []*node { +func (n *node) split(pageSize uintptr) []*node { var nodes []*node node := n @@ -270,7 +266,7 @@ func (n *node) split(pageSize int) []*node { // splitTwo breaks up a node into two smaller nodes, if appropriate. // This should only be called from the split() function. -func (n *node) splitTwo(pageSize int) (*node, *node) { +func (n *node) splitTwo(pageSize uintptr) (*node, *node) { // Ignore the split if the page doesn't have at least enough nodes for // two pages or if the nodes can fit in a single page. if len(n.inodes) <= (minKeysPerPage*2) || n.sizeLessThan(pageSize) { @@ -312,18 +308,18 @@ func (n *node) splitTwo(pageSize int) (*node, *node) { // splitIndex finds the position where a page will fill a given threshold. // It returns the index as well as the size of the first page. // This is only be called from split(). -func (n *node) splitIndex(threshold int) (index, sz int) { +func (n *node) splitIndex(threshold int) (index, sz uintptr) { sz = pageHeaderSize // Loop until we only have the minimum number of keys required for the second page. for i := 0; i < len(n.inodes)-minKeysPerPage; i++ { - index = i + index = uintptr(i) inode := n.inodes[i] - elsize := n.pageElementSize() + len(inode.key) + len(inode.value) + elsize := n.pageElementSize() + uintptr(len(inode.key)) + uintptr(len(inode.value)) // If we have at least the minimum number of keys and adding another // node would put us over the threshold then exit and return. - if i >= minKeysPerPage && sz+elsize > threshold { + if index >= minKeysPerPage && sz+elsize > uintptr(threshold) { break } @@ -356,7 +352,7 @@ func (n *node) spill() error { n.children = nil // Split nodes into appropriate sizes. The first node will always be n. - var nodes = n.split(tx.db.pageSize) + var nodes = n.split(uintptr(tx.db.pageSize)) for _, node := range nodes { // Add node's page to the freelist if it's not new. if node.pgid > 0 { @@ -365,7 +361,7 @@ func (n *node) spill() error { } // Allocate contiguous space for the node. - p, err := tx.allocate((node.size() / tx.db.pageSize) + 1) + p, err := tx.allocate((node.size() + tx.db.pageSize - 1) / tx.db.pageSize) if err != nil { return err } @@ -587,9 +583,11 @@ func (n *node) dump() { type nodes []*node -func (s nodes) Len() int { return len(s) } -func (s nodes) Swap(i, j int) { s[i], s[j] = s[j], s[i] } -func (s nodes) Less(i, j int) bool { return bytes.Compare(s[i].inodes[0].key, s[j].inodes[0].key) == -1 } +func (s nodes) Len() int { return len(s) } +func (s nodes) Swap(i, j int) { s[i], s[j] = s[j], s[i] } +func (s nodes) Less(i, j int) bool { + return bytes.Compare(s[i].inodes[0].key, s[j].inodes[0].key) == -1 +} // inode represents an internal node inside of a node. // It can be used to point to elements in a page or point diff --git a/vendor/github.com/boltdb/bolt/page.go b/vendor/go.etcd.io/bbolt/page.go similarity index 77% rename from vendor/github.com/boltdb/bolt/page.go rename to vendor/go.etcd.io/bbolt/page.go index cde403ae..c9a158fb 100644 --- a/vendor/github.com/boltdb/bolt/page.go +++ b/vendor/go.etcd.io/bbolt/page.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "fmt" @@ -7,12 +7,12 @@ import ( "unsafe" ) -const pageHeaderSize = int(unsafe.Offsetof(((*page)(nil)).ptr)) +const pageHeaderSize = unsafe.Sizeof(page{}) const minKeysPerPage = 2 -const branchPageElementSize = int(unsafe.Sizeof(branchPageElement{})) -const leafPageElementSize = int(unsafe.Sizeof(leafPageElement{})) +const branchPageElementSize = unsafe.Sizeof(branchPageElement{}) +const leafPageElementSize = unsafe.Sizeof(leafPageElement{}) const ( branchPageFlag = 0x01 @@ -32,7 +32,6 @@ type page struct { flags uint16 count uint16 overflow uint32 - ptr uintptr } // typ returns a human readable page type string used for debugging. @@ -51,13 +50,13 @@ func (p *page) typ() string { // meta returns a pointer to the metadata section of the page. func (p *page) meta() *meta { - return (*meta)(unsafe.Pointer(&p.ptr)) + return (*meta)(unsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p))) } // leafPageElement retrieves the leaf node by index func (p *page) leafPageElement(index uint16) *leafPageElement { - n := &((*[0x7FFFFFF]leafPageElement)(unsafe.Pointer(&p.ptr)))[index] - return n + return (*leafPageElement)(unsafeIndex(unsafe.Pointer(p), unsafe.Sizeof(*p), + leafPageElementSize, int(index))) } // leafPageElements retrieves a list of leaf nodes. @@ -65,12 +64,16 @@ func (p *page) leafPageElements() []leafPageElement { if p.count == 0 { return nil } - return ((*[0x7FFFFFF]leafPageElement)(unsafe.Pointer(&p.ptr)))[:] + var elems []leafPageElement + data := unsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p)) + unsafeSlice(unsafe.Pointer(&elems), data, int(p.count)) + return elems } // branchPageElement retrieves the branch node by index func (p *page) branchPageElement(index uint16) *branchPageElement { - return &((*[0x7FFFFFF]branchPageElement)(unsafe.Pointer(&p.ptr)))[index] + return (*branchPageElement)(unsafeIndex(unsafe.Pointer(p), unsafe.Sizeof(*p), + unsafe.Sizeof(branchPageElement{}), int(index))) } // branchPageElements retrieves a list of branch nodes. @@ -78,12 +81,15 @@ func (p *page) branchPageElements() []branchPageElement { if p.count == 0 { return nil } - return ((*[0x7FFFFFF]branchPageElement)(unsafe.Pointer(&p.ptr)))[:] + var elems []branchPageElement + data := unsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p)) + unsafeSlice(unsafe.Pointer(&elems), data, int(p.count)) + return elems } // dump writes n bytes of the page to STDERR as hex output. func (p *page) hexdump(n int) { - buf := (*[maxAllocSize]byte)(unsafe.Pointer(p))[:n] + buf := unsafeByteSlice(unsafe.Pointer(p), 0, 0, n) fmt.Fprintf(os.Stderr, "%x\n", buf) } @@ -102,8 +108,7 @@ type branchPageElement struct { // key returns a byte slice of the node key. func (n *branchPageElement) key() []byte { - buf := (*[maxAllocSize]byte)(unsafe.Pointer(n)) - return (*[maxAllocSize]byte)(unsafe.Pointer(&buf[n.pos]))[:n.ksize] + return unsafeByteSlice(unsafe.Pointer(n), 0, int(n.pos), int(n.pos)+int(n.ksize)) } // leafPageElement represents a node on a leaf page. @@ -116,14 +121,16 @@ type leafPageElement struct { // key returns a byte slice of the node key. func (n *leafPageElement) key() []byte { - buf := (*[maxAllocSize]byte)(unsafe.Pointer(n)) - return (*[maxAllocSize]byte)(unsafe.Pointer(&buf[n.pos]))[:n.ksize:n.ksize] + i := int(n.pos) + j := i + int(n.ksize) + return unsafeByteSlice(unsafe.Pointer(n), 0, i, j) } // value returns a byte slice of the node value. func (n *leafPageElement) value() []byte { - buf := (*[maxAllocSize]byte)(unsafe.Pointer(n)) - return (*[maxAllocSize]byte)(unsafe.Pointer(&buf[n.pos+n.ksize]))[:n.vsize:n.vsize] + i := int(n.pos) + int(n.ksize) + j := i + int(n.vsize) + return unsafeByteSlice(unsafe.Pointer(n), 0, i, j) } // PageInfo represents human readable information about a page. diff --git a/vendor/github.com/boltdb/bolt/tx.go b/vendor/go.etcd.io/bbolt/tx.go similarity index 88% rename from vendor/github.com/boltdb/bolt/tx.go rename to vendor/go.etcd.io/bbolt/tx.go index 6700308a..4b1a64a8 100644 --- a/vendor/github.com/boltdb/bolt/tx.go +++ b/vendor/go.etcd.io/bbolt/tx.go @@ -1,4 +1,4 @@ -package bolt +package bbolt import ( "fmt" @@ -126,10 +126,7 @@ func (tx *Tx) DeleteBucket(name []byte) error { // the error is returned to the caller. func (tx *Tx) ForEach(fn func(name []byte, b *Bucket) error) error { return tx.root.ForEach(func(k, v []byte) error { - if err := fn(k, tx.root.Bucket(k)); err != nil { - return err - } - return nil + return fn(k, tx.root.Bucket(k)) }) } @@ -169,28 +166,18 @@ func (tx *Tx) Commit() error { // Free the old root bucket. tx.meta.root.root = tx.root.root - opgid := tx.meta.pgid - - // Free the freelist and allocate new pages for it. This will overestimate - // the size of the freelist but not underestimate the size (which would be bad). - tx.db.freelist.free(tx.meta.txid, tx.db.page(tx.meta.freelist)) - p, err := tx.allocate((tx.db.freelist.size() / tx.db.pageSize) + 1) - if err != nil { - tx.rollback() - return err + // Free the old freelist because commit writes out a fresh freelist. + if tx.meta.freelist != pgidNoFreelist { + tx.db.freelist.free(tx.meta.txid, tx.db.page(tx.meta.freelist)) } - if err := tx.db.freelist.write(p); err != nil { - tx.rollback() - return err - } - tx.meta.freelist = p.id - // If the high water mark has moved up then attempt to grow the database. - if tx.meta.pgid > opgid { - if err := tx.db.grow(int(tx.meta.pgid+1) * tx.db.pageSize); err != nil { - tx.rollback() + if !tx.db.NoFreelistSync { + err := tx.commitFreelist() + if err != nil { return err } + } else { + tx.meta.freelist = pgidNoFreelist } // Write dirty pages to disk. @@ -235,6 +222,31 @@ func (tx *Tx) Commit() error { return nil } +func (tx *Tx) commitFreelist() error { + // Allocate new pages for the new free list. This will overestimate + // the size of the freelist but not underestimate the size (which would be bad). + opgid := tx.meta.pgid + p, err := tx.allocate((tx.db.freelist.size() / tx.db.pageSize) + 1) + if err != nil { + tx.rollback() + return err + } + if err := tx.db.freelist.write(p); err != nil { + tx.rollback() + return err + } + tx.meta.freelist = p.id + // If the high water mark has moved up then attempt to grow the database. + if tx.meta.pgid > opgid { + if err := tx.db.grow(int(tx.meta.pgid+1) * tx.db.pageSize); err != nil { + tx.rollback() + return err + } + } + + return nil +} + // Rollback closes the transaction and ignores all previous updates. Read-only // transactions must be rolled back and not committed. func (tx *Tx) Rollback() error { @@ -242,17 +254,36 @@ func (tx *Tx) Rollback() error { if tx.db == nil { return ErrTxClosed } - tx.rollback() + tx.nonPhysicalRollback() return nil } +// nonPhysicalRollback is called when user calls Rollback directly, in this case we do not need to reload the free pages from disk. +func (tx *Tx) nonPhysicalRollback() { + if tx.db == nil { + return + } + if tx.writable { + tx.db.freelist.rollback(tx.meta.txid) + } + tx.close() +} + +// rollback needs to reload the free pages from disk in case some system error happens like fsync error. func (tx *Tx) rollback() { if tx.db == nil { return } if tx.writable { tx.db.freelist.rollback(tx.meta.txid) - tx.db.freelist.reload(tx.db.page(tx.db.meta().freelist)) + if !tx.db.hasSyncedFreelist() { + // Reconstruct free page list by scanning the DB to get the whole free page list. + // Note: scaning the whole db is heavy if your db size is large in NoSyncFreeList mode. + tx.db.freelist.noSyncReload(tx.db.freepages()) + } else { + // Read free page list from freelist page. + tx.db.freelist.reload(tx.db.page(tx.db.meta().freelist)) + } } tx.close() } @@ -291,7 +322,9 @@ func (tx *Tx) close() { } // Copy writes the entire database to a writer. -// This function exists for backwards compatibility. Use WriteTo() instead. +// This function exists for backwards compatibility. +// +// Deprecated; Use WriteTo() instead. func (tx *Tx) Copy(w io.Writer) error { _, err := tx.WriteTo(w) return err @@ -301,11 +334,15 @@ func (tx *Tx) Copy(w io.Writer) error { // If err == nil then exactly tx.Size() bytes will be written into the writer. func (tx *Tx) WriteTo(w io.Writer) (n int64, err error) { // Attempt to open reader with WriteFlag - f, err := os.OpenFile(tx.db.path, os.O_RDONLY|tx.WriteFlag, 0) + f, err := tx.db.openFile(tx.db.path, os.O_RDONLY|tx.WriteFlag, 0) if err != nil { return 0, err } - defer func() { _ = f.Close() }() + defer func() { + if cerr := f.Close(); err == nil { + err = cerr + } + }() // Generate a meta page. We use the same page data for both meta pages. buf := make([]byte, tx.db.pageSize) @@ -333,7 +370,7 @@ func (tx *Tx) WriteTo(w io.Writer) (n int64, err error) { } // Move past the meta pages in the file. - if _, err := f.Seek(int64(tx.db.pageSize*2), os.SEEK_SET); err != nil { + if _, err := f.Seek(int64(tx.db.pageSize*2), io.SeekStart); err != nil { return n, fmt.Errorf("seek: %s", err) } @@ -344,14 +381,14 @@ func (tx *Tx) WriteTo(w io.Writer) (n int64, err error) { return n, err } - return n, f.Close() + return n, nil } // CopyFile copies the entire database to file at the given path. // A reader transaction is maintained during the copy so it is safe to continue // using the database while a copy is in progress. func (tx *Tx) CopyFile(path string, mode os.FileMode) error { - f, err := os.OpenFile(path, os.O_RDWR|os.O_CREATE|os.O_TRUNC, mode) + f, err := tx.db.openFile(path, os.O_RDWR|os.O_CREATE|os.O_TRUNC, mode) if err != nil { return err } @@ -379,6 +416,9 @@ func (tx *Tx) Check() <-chan error { } func (tx *Tx) check(ch chan error) { + // Force loading free list if opened in ReadOnly mode. + tx.db.loadFreelist() + // Check if any pages are double freed. freed := make(map[pgid]bool) all := make([]pgid, tx.db.freelist.count()) @@ -394,8 +434,10 @@ func (tx *Tx) check(ch chan error) { reachable := make(map[pgid]*page) reachable[0] = tx.page(0) // meta0 reachable[1] = tx.page(1) // meta1 - for i := uint32(0); i <= tx.page(tx.meta.freelist).overflow; i++ { - reachable[tx.meta.freelist+pgid(i)] = tx.page(tx.meta.freelist) + if tx.meta.freelist != pgidNoFreelist { + for i := uint32(0); i <= tx.page(tx.meta.freelist).overflow; i++ { + reachable[tx.meta.freelist+pgid(i)] = tx.page(tx.meta.freelist) + } } // Recursively check buckets. @@ -453,7 +495,7 @@ func (tx *Tx) checkBucket(b *Bucket, reachable map[pgid]*page, freed map[pgid]bo // allocate returns a contiguous block of memory starting at a given page. func (tx *Tx) allocate(count int) (*page, error) { - p, err := tx.db.allocate(count) + p, err := tx.db.allocate(tx.meta.txid, count) if err != nil { return nil, err } @@ -462,7 +504,7 @@ func (tx *Tx) allocate(count int) (*page, error) { tx.pages[p.id] = p // Update statistics. - tx.stats.PageCount++ + tx.stats.PageCount += count tx.stats.PageAlloc += count * tx.db.pageSize return p, nil @@ -481,20 +523,18 @@ func (tx *Tx) write() error { // Write pages to disk in order. for _, p := range pages { - size := (int(p.overflow) + 1) * tx.db.pageSize + rem := (uint64(p.overflow) + 1) * uint64(tx.db.pageSize) offset := int64(p.id) * int64(tx.db.pageSize) + var written uintptr // Write out page in "max allocation" sized chunks. - ptr := (*[maxAllocSize]byte)(unsafe.Pointer(p)) for { - // Limit our write to our max allocation size. - sz := size + sz := rem if sz > maxAllocSize-1 { sz = maxAllocSize - 1 } + buf := unsafeByteSlice(unsafe.Pointer(p), written, 0, int(sz)) - // Write chunk to disk. - buf := ptr[:sz] if _, err := tx.db.ops.writeAt(buf, offset); err != nil { return err } @@ -503,14 +543,14 @@ func (tx *Tx) write() error { tx.stats.Write++ // Exit inner for loop if we've written all the chunks. - size -= sz - if size == 0 { + rem -= sz + if rem == 0 { break } // Otherwise move offset forward and move pointer to next chunk. offset += int64(sz) - ptr = (*[maxAllocSize]byte)(unsafe.Pointer(&ptr[sz])) + written += uintptr(sz) } } @@ -529,7 +569,7 @@ func (tx *Tx) write() error { continue } - buf := (*[maxAllocSize]byte)(unsafe.Pointer(p))[:tx.db.pageSize] + buf := unsafeByteSlice(unsafe.Pointer(p), 0, 0, tx.db.pageSize) // See https://go.googlesource.com/go/+/f03c9202c43e0abb130669852082117ca50aa9b1 for i := range buf { diff --git a/vendor/go.etcd.io/bbolt/unsafe.go b/vendor/go.etcd.io/bbolt/unsafe.go new file mode 100644 index 00000000..c0e50375 --- /dev/null +++ b/vendor/go.etcd.io/bbolt/unsafe.go @@ -0,0 +1,39 @@ +package bbolt + +import ( + "reflect" + "unsafe" +) + +func unsafeAdd(base unsafe.Pointer, offset uintptr) unsafe.Pointer { + return unsafe.Pointer(uintptr(base) + offset) +} + +func unsafeIndex(base unsafe.Pointer, offset uintptr, elemsz uintptr, n int) unsafe.Pointer { + return unsafe.Pointer(uintptr(base) + offset + uintptr(n)*elemsz) +} + +func unsafeByteSlice(base unsafe.Pointer, offset uintptr, i, j int) []byte { + // See: https://github.com/golang/go/wiki/cgo#turning-c-arrays-into-go-slices + // + // This memory is not allocated from C, but it is unmanaged by Go's + // garbage collector and should behave similarly, and the compiler + // should produce similar code. Note that this conversion allows a + // subslice to begin after the base address, with an optional offset, + // while the URL above does not cover this case and only slices from + // index 0. However, the wiki never says that the address must be to + // the beginning of a C allocation (or even that malloc was used at + // all), so this is believed to be correct. + return (*[maxAllocSize]byte)(unsafeAdd(base, offset))[i:j:j] +} + +// unsafeSlice modifies the data, len, and cap of a slice variable pointed to by +// the slice parameter. This helper should be used over other direct +// manipulation of reflect.SliceHeader to prevent misuse, namely, converting +// from reflect.SliceHeader to a Go slice type. +func unsafeSlice(slice, data unsafe.Pointer, len int) { + s := (*reflect.SliceHeader)(slice) + s.Data = uintptr(data) + s.Cap = len + s.Len = len +} diff --git a/vendor/modules.txt b/vendor/modules.txt index cb60fa79..983fd230 100644 --- a/vendor/modules.txt +++ b/vendor/modules.txt @@ -1,13 +1,11 @@ # cloud.google.com/go v0.74.0 ## explicit cloud.google.com/go/compute/metadata -# github.com/TwinProduction/gocache v1.2.0 +# github.com/TwinProduction/gocache v1.2.1 ## explicit github.com/TwinProduction/gocache # github.com/beorn7/perks v1.0.1 github.com/beorn7/perks/quantile -# github.com/boltdb/bolt v1.3.1 -github.com/boltdb/bolt # github.com/cespare/xxhash/v2 v2.1.1 github.com/cespare/xxhash/v2 # github.com/davecgh/go-spew v1.1.1 @@ -67,6 +65,8 @@ github.com/prometheus/procfs/internal/fs github.com/prometheus/procfs/internal/util # github.com/spf13/pflag v1.0.5 github.com/spf13/pflag +# go.etcd.io/bbolt v1.3.5 +go.etcd.io/bbolt # golang.org/x/crypto v0.0.0-20201221181555-eec23a3978ad ## explicit golang.org/x/crypto/ed25519