opal-up 0.0.2 → 0.0.3

data/ext/up_ext/AsyncSocket.h

@@ -0,0 +1,355 @@
+/*
+ * Authored by Alex Hultman, 2018-2020.
+ * Intellectual property of third-party.
+
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+
+ *     http://www.apache.org/licenses/LICENSE-2.0
+
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#ifndef UWS_ASYNCSOCKET_H
+#define UWS_ASYNCSOCKET_H
+
+/* This class implements async socket memory management strategies */
+
+/* NOTE: Many unsigned/signed conversion warnings could be solved by moving from int length
+ * to unsigned length for everything to/from uSockets - this would however remove the opportunity
+ * to signal error with -1 (which is how the entire UNIX syscalling is built). */
+
+#include <cstring>
+#include <iostream>
+
+#include "libusockets.h"
+
+#include "LoopData.h"
+#include "AsyncSocketData.h"
+
+namespace uWS {
+
+    enum SendBufferAttribute {
+        NEEDS_NOTHING,
+        NEEDS_DRAIN,
+        NEEDS_UNCORK
+    };
+
+    template <bool, bool, typename> struct WebSocketContext;
+
+template <bool SSL>
+struct AsyncSocket {
+    /* This guy is promiscuous */
+    template <bool> friend struct HttpContext;
+    template <bool, bool, typename> friend struct WebSocketContext;
+    template <bool> friend struct TemplatedApp;
+    template <bool, typename> friend struct WebSocketContextData;
+    template <typename, typename> friend struct TopicTree;
+    template <bool> friend struct HttpResponse;
+
+private:
+    /* Helper, do not use directly (todo: move to uSockets or de-crazify) */
+    void throttle_helper(int toggle) {
+        /* These should be exposed by uSockets */
+        static thread_local int us_events[2] = {0, 0};
+
+        struct us_poll_t *p = (struct us_poll_t *) this;
+        struct us_loop_t *loop = us_socket_context_loop(SSL, us_socket_context(SSL, (us_socket_t *) this));
+
+        if (toggle) {
+            /* Pause */
+            int events = us_poll_events(p);
+            if (events) {
+                us_events[getBufferedAmount() ? 1 : 0] = events;
+            }
+            us_poll_change(p, loop, 0);
+        } else {
+            /* Resume */
+            int events = us_events[getBufferedAmount() ? 1 : 0];
+            us_poll_change(p, loop, events);
+        }
+    }
+
+protected:
+    /* Returns SSL pointer or FD as pointer */
+    void *getNativeHandle() {
+        return us_socket_get_native_handle(SSL, (us_socket_t *) this);
+    }
+
+    /* Get loop data for socket */
+    LoopData *getLoopData() {
+        return (LoopData *) us_loop_ext(us_socket_context_loop(SSL, us_socket_context(SSL, (us_socket_t *) this)));
+    }
+
+    /* Get socket extension */
+    AsyncSocketData<SSL> *getAsyncSocketData() {
+        return (AsyncSocketData<SSL> *) us_socket_ext(SSL, (us_socket_t *) this);
+    }
+
+    /* Socket timeout */
+    void timeout(unsigned int seconds) {
+        us_socket_timeout(SSL, (us_socket_t *) this, seconds);
+    }
+
+    /* Shutdown socket without any automatic drainage */
+    void shutdown() {
+        us_socket_shutdown(SSL, (us_socket_t *) this);
+    }
+
+    /* Experimental pause */
+    us_socket_t *pause() {
+        throttle_helper(1);
+        return (us_socket_t *) this;
+    }
+
+    /* Experimental resume */
+    us_socket_t *resume() {
+        throttle_helper(0);
+        return (us_socket_t *) this;
+    }
+
+    /* Immediately close socket */
+    us_socket_t *close() {
+        return us_socket_close(SSL, (us_socket_t *) this, 0, nullptr);
+    }
+
+    void corkUnchecked() {
+        /* What if another socket is corked? */
+        getLoopData()->corkedSocket = this;
+    }
+
+    void uncorkWithoutSending() {
+        if (isCorked()) {
+            getLoopData()->corkedSocket = nullptr;
+        }
+    }
+
+    /* Cork this socket. Only one socket may ever be corked per-loop at any given time */
+    void cork() {
+        /* Extra check for invalid corking of others */
+        if (getLoopData()->corkOffset && getLoopData()->corkedSocket != this) {
+            std::cerr << "Error: Cork buffer must not be acquired without checking canCork!" << std::endl;
+            std::terminate();
+        }
+
+        /* What if another socket is corked? */
+        getLoopData()->corkedSocket = this;
+    }
+
+    /* Returns wheter we are corked or not */
+    bool isCorked() {
+        return getLoopData()->corkedSocket == this;
+    }
+
+    /* Returns whether we could cork (it is free) */
+    bool canCork() {
+        return getLoopData()->corkedSocket == nullptr;
+    }
+
+    /* Returns a suitable buffer for temporary assemblation of send data */
+    std::pair<char *, SendBufferAttribute> getSendBuffer(size_t size) {
+        /* First step is to determine if we already have backpressure or not */
+        LoopData *loopData = getLoopData();
+        BackPressure &backPressure = getAsyncSocketData()->buffer;
+        size_t existingBackpressure = backPressure.length();
+        if ((!existingBackpressure) && (isCorked() || canCork()) && (loopData->corkOffset + size < LoopData::CORK_BUFFER_SIZE)) {
+            /* Cork automatically if we can */
+            if (isCorked()) {
+                char *sendBuffer = loopData->corkBuffer + loopData->corkOffset;
+                loopData->corkOffset += (unsigned int) size;
+                return {sendBuffer, SendBufferAttribute::NEEDS_NOTHING};
+            } else {
+                cork();
+                char *sendBuffer = loopData->corkBuffer + loopData->corkOffset;
+                loopData->corkOffset += (unsigned int) size;
+                return {sendBuffer, SendBufferAttribute::NEEDS_UNCORK};
+            }
+        } else {
+
+            /* If we are corked and there is already data in the cork buffer,
+            mark how much is ours and reset it */
+            unsigned int ourCorkOffset = 0;
+            if (isCorked() && loopData->corkOffset) {
+                ourCorkOffset = loopData->corkOffset;
+                loopData->corkOffset = 0;
+            }
+
+            /* Fallback is to use the backpressure as buffer */
+            backPressure.resize(ourCorkOffset + existingBackpressure + size);
+
+            /* And copy corkbuffer in front */
+            memcpy((char *) backPressure.data() + existingBackpressure, loopData->corkBuffer, ourCorkOffset);
+
+            return {(char *) backPressure.data() + ourCorkOffset + existingBackpressure, SendBufferAttribute::NEEDS_DRAIN};
+        }
+    }
+
+    /* Returns the user space backpressure. */
+    unsigned int getBufferedAmount() {
+        /* We return the actual amount of bytes in backbuffer, including pendingRemoval */
+        return (unsigned int) getAsyncSocketData()->buffer.totalLength();
+    }
+
+    /* Returns the text representation of an IPv4 or IPv6 address */
+    std::string_view addressAsText(std::string_view binary) {
+        static thread_local char buf[64];
+        int ipLength = 0;
+
+        if (!binary.length()) {
+            return {};
+        }
+
+        unsigned char *b = (unsigned char *) binary.data();
+
+        if (binary.length() == 4) {
+            ipLength = snprintf(buf, 64, "%u.%u.%u.%u", b[0], b[1], b[2], b[3]);
+        } else {
+            ipLength = snprintf(buf, 64, "%02x%02x:%02x%02x:%02x%02x:%02x%02x:%02x%02x:%02x%02x:%02x%02x:%02x%02x",
+                b[0], b[1], b[2], b[3], b[4], b[5], b[6], b[7], b[8], b[9], b[10], b[11],
+                b[12], b[13], b[14], b[15]);
+        }
+
+        return {buf, (unsigned int) ipLength};
+    }
+
+    /* Returns the remote IP address or empty string on failure */
+    std::string_view getRemoteAddress() {
+        static thread_local char buf[16];
+        int ipLength = 16;
+        us_socket_remote_address(SSL, (us_socket_t *) this, buf, &ipLength);
+        return std::string_view(buf, (unsigned int) ipLength);
+    }
+
+    /* Returns the text representation of IP */
+    std::string_view getRemoteAddressAsText() {
+        return addressAsText(getRemoteAddress());
+    }
+
+    /* Write in three levels of prioritization: cork-buffer, syscall, socket-buffer. Always drain if possible.
+     * Returns pair of bytes written (anywhere) and wheter or not this call resulted in the polling for
+     * writable (or we are in a state that implies polling for writable). */
+    std::pair<int, bool> write(const char *src, int length, bool optionally = false, int nextLength = 0) {
+        /* Fake success if closed, simple fix to allow uncork of closed socket to succeed */
+        if (us_socket_is_closed(SSL, (us_socket_t *) this)) {
+            return {length, false};
+        }
+
+        LoopData *loopData = getLoopData();
+        AsyncSocketData<SSL> *asyncSocketData = getAsyncSocketData();
+
+        /* We are limited if we have a per-socket buffer */
+        if (asyncSocketData->buffer.length()) {
+            /* Write off as much as we can */
+            int written = us_socket_write(SSL, (us_socket_t *) this, asyncSocketData->buffer.data(), (int) asyncSocketData->buffer.length(), /*nextLength != 0 | */length);
+
+            /* On failure return, otherwise continue down the function */
+            if ((unsigned int) written < asyncSocketData->buffer.length()) {
+
+                /* Update buffering (todo: we can do better here if we keep track of what happens to this guy later on) */
+                asyncSocketData->buffer.erase((unsigned int) written);
+
+                if (optionally) {
+                    /* Thankfully we can exit early here */
+                    return {0, true};
+                } else {
+                    /* This path is horrible and points towards erroneous usage */
+                    asyncSocketData->buffer.append(src, (unsigned int) length);
+
+                    return {length, true};
+                }
+            }
+
+            /* At this point we simply have no buffer and can continue as normal */
+            asyncSocketData->buffer.clear();
+        }
+
+        if (length) {
+            if (loopData->corkedSocket == this) {
+                /* We are corked */
+                if (LoopData::CORK_BUFFER_SIZE - loopData->corkOffset >= (unsigned int) length) {
+                    /* If the entire chunk fits in cork buffer */
+                    memcpy(loopData->corkBuffer + loopData->corkOffset, src, (unsigned int) length);
+                    loopData->corkOffset += (unsigned int) length;
+                    /* Fall through to default return */
+                } else {
+                    /* Strategy differences between SSL and non-SSL regarding syscall minimizing */
+                    if constexpr (false) {
+                        /* Cork up as much as we can */
+                        unsigned int stripped = LoopData::CORK_BUFFER_SIZE - loopData->corkOffset;
+                        memcpy(loopData->corkBuffer + loopData->corkOffset, src, stripped);
+                        loopData->corkOffset = LoopData::CORK_BUFFER_SIZE;
+
+                        auto [written, failed] = uncork(src + stripped, length - (int) stripped, optionally);
+                        return {written + (int) stripped, failed};
+                    }
+
+                    /* For non-SSL we take the penalty of two syscalls */
+                    return uncork(src, length, optionally);
+                }
+            } else {
+                /* We are not corked */
+                int written = us_socket_write(SSL, (us_socket_t *) this, src, length, nextLength != 0);
+
+                /* Did we fail? */
+                if (written < length) {
+                    /* If the write was optional then just bail out */
+                    if (optionally) {
+                        return {written, true};
+                    }
+
+                    /* Fall back to worst possible case (should be very rare for HTTP) */
+                    /* At least we can reserve room for next chunk if we know it up front */
+                    if (nextLength) {
+                        asyncSocketData->buffer.reserve(asyncSocketData->buffer.length() + (size_t) (length - written + nextLength));
+                    }
+
+                    /* Buffer this chunk */
+                    asyncSocketData->buffer.append(src + written, (size_t) (length - written));
+
+                    /* Return the failure */
+                    return {length, true};
+                }
+                /* Fall through to default return */
+            }
+        }
+
+        /* Default fall through return */
+        return {length, false};
+    }
+
+    /* Uncork this socket and flush or buffer any corked and/or passed data. It is essential to remember doing this. */
+    /* It does NOT count bytes written from cork buffer (they are already accounted for in the write call responsible for its corking)! */
+    std::pair<int, bool> uncork(const char *src = nullptr, int length = 0, bool optionally = false) {
+        LoopData *loopData = getLoopData();
+
+        if (loopData->corkedSocket == this) {
+            loopData->corkedSocket = nullptr;
+
+            if (loopData->corkOffset) {
+                /* Corked data is already accounted for via its write call */
+                auto [written, failed] = write(loopData->corkBuffer, (int) loopData->corkOffset, false, length);
+                loopData->corkOffset = 0;
+
+                if (failed) {
+                    /* We do not need to care for buffering here, write does that */
+                    return {0, true};
+                }
+            }
+
+            /* We should only return with new writes, not things written to cork already */
+            return write(src, length, optionally, 0);
+        } else {
+            /* We are not even corked! */
+            return {0, false};
+        }
+    }
+};
+
+}
+
+#endif // UWS_ASYNCSOCKET_H

data/ext/up_ext/AsyncSocketData.h

@@ -0,0 +1,87 @@
+/*
+ * Authored by Alex Hultman, 2018-2021.
+ * Intellectual property of third-party.
+
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+
+ *     http://www.apache.org/licenses/LICENSE-2.0
+
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#ifndef UWS_ASYNCSOCKETDATA_H
+#define UWS_ASYNCSOCKETDATA_H
+
+#include <string>
+
+namespace uWS {
+
+struct BackPressure {
+    std::string buffer;
+    unsigned int pendingRemoval = 0;
+    BackPressure(BackPressure &&other) {
+        buffer = std::move(other.buffer);
+        pendingRemoval = other.pendingRemoval;
+    }
+    BackPressure() = default;
+    void append(const char *data, size_t length) {
+        buffer.append(data, length);
+    }
+    void erase(unsigned int length) {
+        pendingRemoval += length;
+        /* Always erase a minimum of 1/32th the current backpressure */
+        if (pendingRemoval > (buffer.length() >> 5)) {
+            buffer.erase(0, pendingRemoval);
+            pendingRemoval = 0;
+        }
+    }
+    size_t length() {
+        return buffer.length() - pendingRemoval;
+    }
+    void clear() {
+        pendingRemoval = 0;
+        buffer.clear();
+    }
+    void reserve(size_t length) {
+        buffer.reserve(length + pendingRemoval);
+    }
+    void resize(size_t length) {
+        buffer.resize(length + pendingRemoval);
+    }
+    const char *data() {
+        return buffer.data() + pendingRemoval;
+    }
+    size_t size() {
+        return length();
+    }
+    /* The total length, incuding pending removal */
+    size_t totalLength() {
+        return buffer.length();
+    }
+};
+
+/* Depending on how we want AsyncSocket to function, this will need to change */
+
+template <bool SSL>
+struct AsyncSocketData {
+    /* This will do for now */
+    BackPressure buffer;
+
+    /* Allow move constructing us */
+    AsyncSocketData(BackPressure &&backpressure) : buffer(std::move(backpressure)) {
+
+    }
+
+    /* Or emppty */
+    AsyncSocketData() = default;
+};
+
+}
+
+#endif // UWS_ASYNCSOCKETDATA_H

data/ext/up_ext/BloomFilter.h

@@ -0,0 +1,83 @@
+/*
+ * Authored by Alex Hultman, 2018-2022.
+ * Intellectual property of third-party.
+
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+
+ *     http://www.apache.org/licenses/LICENSE-2.0
+
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#ifndef UWS_BLOOMFILTER_H
+#define UWS_BLOOMFILTER_H
+
+/* This filter has no false positives or collisions for the standard
+ * and non-standard common request headers */
+
+#include <cstdint>
+#include <string_view>
+#include <bitset>
+
+namespace uWS {
+
+struct BloomFilter {
+private:
+    std::bitset<256> filter;
+    static inline uint32_t perfectHash(uint32_t features) {
+        return features *= 1843993368;
+    }
+
+    union ScrambleArea {
+        unsigned char p[4];
+        uint32_t val;
+    };
+
+    ScrambleArea getFeatures(std::string_view key) {
+        ScrambleArea s;
+        s.p[0] = reinterpret_cast<const unsigned char&>(key[0]);
+        s.p[1] = reinterpret_cast<const unsigned char&>(key[key.length() - 1]);
+        s.p[2] = reinterpret_cast<const unsigned char&>(key[key.length() - 2]);
+        s.p[3] = reinterpret_cast<const unsigned char&>(key[key.length() >> 1]);
+        return s;
+    }
+
+public:
+    bool mightHave(std::string_view key) {
+        if (key.length() < 2) {
+            return true;
+        }
+    
+        ScrambleArea s = getFeatures(key);
+        s.val = perfectHash(s.val);
+        return filter[s.p[0]] &&
+        filter[s.p[1]] &&
+        filter[s.p[2]] &&
+        filter[s.p[3]];
+    }
+
+    void add(std::string_view key) {
+        if (key.length() >= 2) {
+            ScrambleArea s = getFeatures(key);
+            s.val = perfectHash(s.val);
+            filter[s.p[0]] = 1;
+            filter[s.p[1]] = 1;
+            filter[s.p[2]] = 1;
+            filter[s.p[3]] = 1;
+        }
+    }
+
+    void reset() {
+        filter.reset();
+    }
+};
+
+}
+
+#endif // UWS_BLOOMFILTER_H