2009-03-03 |
phk | Flexelint nitpickery.
|
commit | commitdiff | tree |
2009-03-03 |
phk | Add minimal facility for dealing with TELNET option...
|
commit | commitdiff | tree |
2009-03-03 |
phk | Add support for authenticating CLI telnet connections
|
commit | commitdiff | tree |
2009-03-03 |
phk | Minor bit of cleanup: initialize CLI from main() instead...
|
commit | commitdiff | tree |
2009-03-02 |
phk | Be much more careful about volatile and locking
|
commit | commitdiff | tree |
2009-03-02 |
phk | I was confused, always rush when we deref an *object...
|
commit | commitdiff | tree |
2009-03-02 |
phk | Add a missing refcnt in the case where we lost a collision...
|
commit | commitdiff | tree |
2009-03-02 |
phk | Clarify the waitinglist handling a little bit
|
commit | commitdiff | tree |
2009-03-02 |
phk | Cache the vcl_fetch{} result, in case vcl_discard{...
|
commit | commitdiff | tree |
2009-03-02 |
phk | Give coords for WRONG
|
commit | commitdiff | tree |
2009-03-02 |
phk | Add a debug facility for substituting digests with...
|
commit | commitdiff | tree |
2009-03-02 |
phk | Add an assert to make sure 32 stays 32
|
commit | commitdiff | tree |
2009-02-28 |
phk | Update this test to not rely of a known dynamic of...
|
commit | commitdiff | tree |
2009-02-28 |
phk | Postpone allocating the actual object until we have...
|
commit | commitdiff | tree |
2009-02-28 |
phk | Use the right workspace for beresp.status
|
commit | commitdiff | tree |
2009-02-28 |
phk | When we know that an object will not make it into the...
|
commit | commitdiff | tree |
2009-02-28 |
phk | Now that we do not need the new object while holding...
|
commit | commitdiff | tree |
2009-02-28 |
phk | On a cache-miss, insert only the (busy) objcore, leave...
|
commit | commitdiff | tree |
2009-02-28 |
phk | Eliminate the parent-child linkage between objects...
|
commit | commitdiff | tree |
2009-02-27 |
phk | Now that ESI parsing is not done while we execute vcl_fetch...
|
commit | commitdiff | tree |
2009-02-27 |
phk | TIMBER!
|
commit | commitdiff | tree |
2009-02-27 |
phk | Change assert to error
|
commit | commitdiff | tree |
2009-02-27 |
phk | Move the http_conn into the bereq.
|
commit | commitdiff | tree |
2009-02-27 |
phk | move object twiddling from FetchHdr() to cache_center.c
|
commit | commitdiff | tree |
2009-02-27 |
phk | Add a beresp.* variable family, which contains the...
|
commit | commitdiff | tree |
2009-02-27 |
phk | Don't pollute bereq.* in vcl_fetch{} with the reply
|
commit | commitdiff | tree |
2009-02-27 |
phk | Sorry, forgot to commit this file.
|
commit | commitdiff | tree |
2009-02-26 |
phk | Add the two new stevedore entrypoints:
|
commit | commitdiff | tree |
2009-02-26 |
phk | Add a alloc/free function.
|
commit | commitdiff | tree |
2009-02-23 |
phk | Explain what functions do.
|
commit | commitdiff | tree |
2009-02-23 |
phk | Read and Write the segment list.
|
commit | commitdiff | tree |
2009-02-23 |
phk | get close to reading segment table
|
commit | commitdiff | tree |
2009-02-23 |
phk | Move on to the worker process side of persistent storage
|
commit | commitdiff | tree |
2009-02-23 |
phk | The SHA256 is outside the signature struct
|
commit | commitdiff | tree |
2009-02-23 |
phk | A bit of FlexeLint nitpicking
|
commit | commitdiff | tree |
2009-02-23 |
phk | Create/validate the three signed tables (ban1+2, segments)
|
commit | commitdiff | tree |
2009-02-23 |
phk | Add symbolic SHA256_LEN for the size of the digest.
|
commit | commitdiff | tree |
2009-02-23 |
phk | Code to create, map and ident the storage silo
|
commit | commitdiff | tree |
2009-02-22 |
phk | Bring in the first inkling of the persistent storage...
|
commit | commitdiff | tree |
2009-02-22 |
phk | Move the catalogs of storage and hash modules closer...
|
commit | commitdiff | tree |
2009-02-20 |
phk | Some constifications
|
commit | commitdiff | tree |
2009-02-20 |
phk | Handle worker thread on subrequests correctly
|
commit | commitdiff | tree |
2009-02-20 |
phk | Don't clean on every request, that was for debugging...
|
commit | commitdiff | tree |
2009-02-20 |
phk | Don't try to rush waiting sessions of refcount is zero...
|
commit | commitdiff | tree |
2009-02-20 |
phk | Add a lot more paranoia asserts
|
commit | commitdiff | tree |
2009-02-20 |
phk | Put the cleanup next to the preallocation and remember...
|
commit | commitdiff | tree |
2009-02-20 |
phk | Add a delta-stats structure to worker threads and keep
|
commit | commitdiff | tree |
2009-02-20 |
phk | Make "hcb.dump" a CLI command
|
commit | commitdiff | tree |
2009-02-20 |
phk | Add a argument to the stat_field macro, to be used...
|
commit | commitdiff | tree |
2009-02-20 |
phk | Be more defensive around objhead retirement.
|
commit | commitdiff | tree |
2009-02-20 |
phk | Remove unused hoh_digest
|
commit | commitdiff | tree |
2009-02-19 |
phk | Split a couple of generally useful routines out of...
|
commit | commitdiff | tree |
2009-02-17 |
phk | with the prefetch timer out of the way, we do not need...
|
commit | commitdiff | tree |
2009-02-17 |
phk | Now that we only have one timer firing, remove stuff...
|
commit | commitdiff | tree |
2009-02-17 |
phk | Eliminate prefetch, grace was a much better idea.
|
commit | commitdiff | tree |
2009-02-17 |
phk | Assert that we got an objcore
|
commit | commitdiff | tree |
2009-02-17 |
phk | Move the ttl from object up to the objcore
|
commit | commitdiff | tree |
2009-02-17 |
phk | Move the pass flag from the object to the objcore.
|
commit | commitdiff | tree |
2009-02-17 |
phk | Convert assert to error report
|
commit | commitdiff | tree |
2009-02-16 |
phk | Allow for pass opbjects not having a objcore
|
commit | commitdiff | tree |
2009-02-16 |
phk | Move the object busy flag up to the objcore.
|
commit | commitdiff | tree |
2009-02-16 |
phk | Compress objcore a bit more by squezing 32 bit fields...
|
commit | commitdiff | tree |
2009-02-16 |
phk | Reduce objhead size by overloading.
|
commit | commitdiff | tree |
2009-02-13 |
phk | To prevent masking bugs, disable auto_restart by default.
|
commit | commitdiff | tree |
2009-02-13 |
phk | Make sure to clean the VBE properly before releasing it.
|
commit | commitdiff | tree |
2009-02-13 |
phk | Add two hidden CLI commands for panic'ing the master...
|
commit | commitdiff | tree |
2009-02-13 |
phk | Use semaphores instead of delays
|
commit | commitdiff | tree |
2009-02-13 |
phk | Don't copy the debug flag around, make it a global...
|
commit | commitdiff | tree |
2009-02-13 |
phk | vi(1) 'x' key misfire.
|
commit | commitdiff | tree |
2009-02-13 |
phk | Disable CLI banner when running tests.
|
commit | commitdiff | tree |
2009-02-13 |
phk | Add a "banner" command to the CLI interface.
|
commit | commitdiff | tree |
2009-02-13 |
phk | Update copyright year
|
commit | commitdiff | tree |
2009-02-13 |
phk | Add VLU_Data() function for injecting input
|
commit | commitdiff | tree |
2009-02-13 |
phk | parameter name change "acceptor" -> "waiter"
|
commit | commitdiff | tree |
2009-02-13 |
phk | Due to lack og foresight,the threads that keep an eye...
|
commit | commitdiff | tree |
2009-02-12 |
phk | Move the LRU timestamp from the objcore to the object...
|
commit | commitdiff | tree |
2009-02-12 |
phk | Move objcore over objects, so that objhead has a list...
|
commit | commitdiff | tree |
2009-02-11 |
phk | Pull struct objexp out of the closet and call it objcore.
|
commit | commitdiff | tree |
2009-02-03 |
phk | Regression testcase for #433
|
commit | commitdiff | tree |
2009-02-03 |
phk | Overhaul the ESI parser in light of #433 and the prospect...
|
commit | commitdiff | tree |
2009-02-03 |
phk | Add line numbers to this test, so we can see where...
|
commit | commitdiff | tree |
2009-02-03 |
phk | Add a counter for esi parse errors and test it in all...
|
commit | commitdiff | tree |
2009-02-03 |
phk | Add bodylen expects to the ESI test cases
|
commit | commitdiff | tree |
2009-02-03 |
phk | Update this test to actually test what it should.
|
commit | commitdiff | tree |
2009-02-03 |
phk | Make it possible (at great performance loss) to force...
|
commit | commitdiff | tree |
2009-02-03 |
phk | Mark that we do limited segment size fetches for debugging
|
commit | commitdiff | tree |
2009-02-03 |
phk | Opps, debugging hack leaked in.
|
commit | commitdiff | tree |
2009-02-03 |
phk | The esi detector should also spot esi comments.
|
commit | commitdiff | tree |
2009-02-03 |
phk | Add an (unlocked) counter for number of ESI objects...
|
commit | commitdiff | tree |
2009-02-02 |
phk | Close a race where VCL tries to modify the obj.ttl...
|
commit | commitdiff | tree |
2009-01-28 |
phk | More FlexeLint silencing.
|
commit | commitdiff | tree |
2009-01-28 |
phk | silence a couple of Flexelint complaints.
|
commit | commitdiff | tree |
2009-01-27 |
phk | Stop processing ESI elements as soon as we discover...
|
commit | commitdiff | tree |
2009-01-27 |
phk | proper errorchecks for regsub syntax.
|
commit | commitdiff | tree |
2009-01-27 |
phk | Don't modify the obj_timer_when field outside the binheap...
|
commit | commitdiff | tree |
2009-01-27 |
phk | Lack of a backend name is no reasonable cause for a...
|
commit | commitdiff | tree |
2009-01-24 |
phk | Test that we know the purge variable when it is compiled in.
|
commit | commitdiff | tree |
2009-01-24 |
phk | Implement the new-purge case where the entire expression...
|
commit | commitdiff | tree |
2009-01-23 |
phk | First cut at VCL support for new-purge.
|
commit | commitdiff | tree |
2009-01-23 |
phk | Don't append a CRNL to the body.
|
commit | commitdiff | tree |
next |