RFC: use some in-memory state in SimpleCache to quickly cache-miss some CantConditionalize cases

net/http/http_cache_transaction.cc

 // Copyright (c) 2012 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "net/http/http_cache_transaction.h"
 
 #include "build/build_config.h"  // For OS_POSIX
 
 #if defined(OS_POSIX)
 #include <unistd.h>
@@ skipping 2418 matching lines @@
       stale_entry_freshness_ = lifetimes.freshness;
       stale_entry_age_ = response_.headers->GetCurrentAge(
           response_.request_time, response_.response_time,
           cache_->clock_->Now());
     }
   }
 
   return validation_required_by_headers;
 }
 
+bool HttpCache::Transaction::ResponseConditionalizable(
+    std::string* etag_value,
+    std::string* last_modified_value) {
+  DCHECK(response_.headers.get());
+
+  // This only makes sense for cached 200 or 206 responses.
+  if (response_.headers->response_code() != 200 &&
+      response_.headers->response_code() != 206) {
+    return false;
+  }
+
+  // ### this looks scary.
+  DCHECK(response_.headers->response_code() != 206 ||
+         response_.headers->HasStrongValidators());
+
+  // Just use the first available ETag and/or Last-Modified header value.
+  // TODO(darin): Or should we use the last?
+
+  if (response_.headers->GetHttpVersion() >= HttpVersion(1, 1))
+    response_.headers->EnumerateHeader(NULL, "etag", etag_value);
+
+  response_.headers->EnumerateHeader(NULL, "last-modified",
+                                     last_modified_value);
+
+  if (etag_value->empty() && last_modified_value->empty())
+    return false;
+
+  return true;
+}
+
 bool HttpCache::Transaction::ConditionalizeRequest() {
   DCHECK(response_.headers.get());
 
   if (method_ == "PUT" || method_ == "DELETE")
     return false;
 
-  // This only makes sense for cached 200 or 206 responses.
-  if (response_.headers->response_code() != 200 &&
-      response_.headers->response_code() != 206) {
-    return false;
-  }
-
   if (fail_conditionalization_for_test_)
     return false;
 
-  DCHECK(response_.headers->response_code() != 206 ||
-         response_.headers->HasStrongValidators());
+  std::string etag_value;
+  std::string last_modified_value;
+  if (!ResponseConditionalizable(&etag_value, &last_modified_value))
+    return false;
 
-  // Just use the first available ETag and/or Last-Modified header value.
-  // TODO(darin): Or should we use the last?
-
-  std::string etag_value;
-  if (response_.headers->GetHttpVersion() >= HttpVersion(1, 1))
-    response_.headers->EnumerateHeader(NULL, "etag", &etag_value);
-
-  std::string last_modified_value;
-  if (!vary_mismatch_) {
-    response_.headers->EnumerateHeader(NULL, "last-modified",
-                                       &last_modified_value);
+  if (vary_mismatch_) {
+    // Can't rely on last-modified if vary is different.
+    last_modified_value.clear();
+    if (etag_value.empty())
+      return false;
   }
 
-  if (etag_value.empty() && last_modified_value.empty())
-    return false;
-
   if (!partial_) {
     // Need to customize the request, so this forces us to allocate :(
     custom_request_.reset(new HttpRequestInfo(*request_));
     request_ = custom_request_.get();
   }
   DCHECK(custom_request_.get());
 
   bool use_if_range =
       partial_ && !partial_->IsCurrentRangeCached() && !invalid_range_;
 
@@ skipping 19 matching lines @@
           HttpRequestHeaders::kIfRange, last_modified_value);
     } else {
       custom_request_->extra_headers.SetHeader(
           HttpRequestHeaders::kIfModifiedSince, last_modified_value);
     }
   }
 
   return true;
 }
 
+bool HttpCache::Transaction::MaybeRejectBasedOnMemoryEntryData(
+    uint8_t in_memory_info) {
+  // Not going to be clever with those... though is it even set here?
+  if (partial_)
+    return false;
+
+  // We can tell that some entries in cache may be unusable under present
+  // circumstances if:
+  // 1) We aren't using everything because of LOAD_SKIP_CACHE_VALIDATION.
+  // 2) We aren't holding on to results of an unused prefetch.
+  // 3) The entry has zero freshness (so it's known to be expired!)
+  // 4) We know the entry can't be revalidated
+  //    (we may be able to match more things here).
+  if (effective_load_flags_ & LOAD_SKIP_CACHE_VALIDATION)
+    return false;
+
+  if (in_memory_info & HttpCache::OBH_UNUSED_SINCE_PREFETCH)
+    return false;
+
+  if (!(in_memory_info & HttpCache::OBH_ZERO_LIFETIME))
+    return false;
+
+  if (!(in_memory_info & OBH_RESPONSE_CANT_CONDITIONALIZE))
+    return false;
+
+  cache_pending_ = false;
+  couldnt_conditionalize_request_ = true;
+  response_.was_cached = true;
+  UpdateCacheEntryStatus(CacheEntryStatus::ENTRY_CANT_CONDITIONALIZE);
+  // ### ugh, this is bad, but just sketching.
+  // TransitionToState(STATE_SEND_REQUEST);
+  next_state_ = STATE_SEND_REQUEST;

[Maks Orlovich, 2017/06/14 19:59:29]

So I just realized that this implementation is bogus, since we want to go
through CREATE_ENTRY and the like. The previous version didn't have this issue,
though.

+
+  return true;
+}
+
 // We just received some headers from the server. We may have asked for a range,
 // in which case partial_ has an object. This could be the first network request
 // we make to fulfill the original request, or we may be already reading (from
 // the net and / or the cache). If we are not expecting a certain response, we
 // just bypass the cache for this request (but again, maybe we are reading), and
 // delete partial_ (so we are not able to "fix" the headers that we return to
 // the user). This results in either a weird response for the caller (we don't
 // expect it after all), or maybe a range that was not exactly what it was asked
 // for.
 //
@@ skipping 202 matching lines @@
   if (truncated)
     DCHECK_EQ(200, response_.headers->response_code());
 
   // When writing headers, we normally only write the non-transient headers.
   bool skip_transient_headers = true;
   scoped_refptr<PickledIOBuffer> data(new PickledIOBuffer());
   response_.Persist(data->pickle(), skip_transient_headers, truncated);
   data->Done();
 
   io_buf_len_ = data->pickle()->size();
+
+  uint8_t hints = 0;
+  // ### recomputing this too much?
+  if (response_.headers->GetFreshnessLifetimes(response_.response_time)
+          .freshness.is_zero())
+    hints |= HttpCache::OBH_ZERO_LIFETIME;
+  std::string etag_ignored, last_modified_ignored;
+  if (!ResponseConditionalizable(&etag_ignored, &last_modified_ignored))
+    hints |= HttpCache::OBH_RESPONSE_CANT_CONDITIONALIZE;
+  if (response_.unused_since_prefetch)
+    hints |= HttpCache::OBH_UNUSED_SINCE_PREFETCH;
+  cache_->GetCurrentBackend()->SetMemoryEntryData(cache_key_, hints);
   return entry_->disk_entry->WriteData(kResponseInfoIndex, 0, data.get(),
                                        io_buf_len_, io_callback_, true);
 }
 
 int HttpCache::Transaction::OnWriteResponseInfoToEntryComplete(int result) {
   if (!entry_)
     return OK;
   if (net_log_.IsCapturing()) {
     net_log_.EndEventWithNetErrorCode(NetLogEventType::HTTP_CACHE_WRITE_INFO,
                                       result);
@@ skipping 380 matching lines @@
 }
 
 void HttpCache::Transaction::TransitionToState(State state) {
   // Ensure that the state is only set once per Do* state.
   DCHECK(in_do_loop_);
   DCHECK_EQ(STATE_UNSET, next_state_) << "Next state is " << state;
   next_state_ = state;
 }
 
 }  // namespace net