March 2012 Archives by date

• Messages sorted by: [ thread ] [ subject ] [ author ]
• More info on this list...

Starting: Sun Mar 4 15:48:53 UTC 2012
Ending: Thu Mar 29 09:46:09 UTC 2012
Messages: 59

• [master] 86e85bb 3.0 syntax   Andreas Plesner Jacobsen
• [master] 5c9d66c Really change to 3.0 syntax   Andreas Plesner Jacobsen
• [master] 5cd69e0 Eliminate obj arg to FetchBody()   Poul-Henning Kamp
• [master] b399a62 Add a set of VFP method functions which can do various sanity asserts for us in a single place.   Poul-Henning Kamp
• [master] f9e413c Don't assert if we fail to get storage in VFP_Begin()   Poul-Henning Kamp
• [master] 501af0b Add a bo->state that tells us what's going on. Improve error reporting at the same time.   Poul-Henning Kamp
• [master] 071d905 Eliminate the returnvalue from FetchBody(), rely on bo->state instead   Poul-Henning Kamp
• [master] 8f2d2e7 Make FetchBody() pool-task compatible.   Poul-Henning Kamp
• [master] af587ef Grammar in varnishncsa man page   Tollef Fog Heen
• [master] fe828b4 Fixing a typo in the documentation.   Tollef Fog Heen
• [master] 4fe672c Add section on device detection   Lasse Karstensen
• [master] 72cb9f6 typos, change example3 to allow for easier purging   Lasse Karstensen
• [master] 4b6e84e Merge branch 'master' of /home/lkarsten/work/varnish-cache-fork   Lasse Karstensen
• [master] e725920 Remove old Date: header before adding our new one.   Poul-Henning Kamp
• [master] 59c6ae4 Join struct vbo and struct busyobj   Poul-Henning Kamp
• [master] d1d7134 Always try to set another thread on the body-fetch job.   Poul-Henning Kamp
• [master] c5e141b Correct function name   Andreas Plesner Jacobsen
• [master] f740e80 Small doc fixes   Andreas Plesner Jacobsen
• [master] b8d4eb8 Hand FetchBody() its own busyobj->refcount, and have it release it when done.   Poul-Henning Kamp
• [master] b78280f Decrement the n_waitinglist counter when freeing waitinglists.   Martin Blix Grydeland
• [master] 40d62bd Have FetchBody reset bo->vfp once it's done with it.   Poul-Henning Kamp
• [master] 7b25530 These testcases depend on us not streaming, so they get a 503 for the failure to fetch a body. Make them explicitly disable streaming.   Poul-Henning Kamp
• [master] d96e763 Include device class when hashing   Lasse Karstensen
• [master] 9831187 Merge branch 'master' of ssh://git.varnish-cache.org/git/varnish-cache   Lasse Karstensen
• [master] 0326c20 Quick polish over HSH_Lookup():   Poul-Henning Kamp
• [master] aea8559 Merge branch 'master' of ssh://git.varnish-cache.org/git/varnish-cache   Poul-Henning Kamp
• [master] e556146 More deref-reduction:   Poul-Henning Kamp
• [master] c92ad59 Utilize the oc->priv2 variable to store the pointer to the allocating stevedore for the default case (malloc and file)   Martin Blix Grydeland
• [master] 88298f1 More sp -> req argument weakening   Poul-Henning Kamp
• [master] cf81441 Tag oc's with COMPLETE or FAILURE when we fetch the body.   Poul-Henning Kamp
• [master] 0041032 Let the body-fetching thread Unbusy the object.   Poul-Henning Kamp
• [master] 2b21621 Also move dropping failed objects to FetchBody()   Poul-Henning Kamp
• [master] d74bc62 Add a mutex timing flag to diag_bitmap   Martin Blix Grydeland
• [master] 6393f26 Fix a misfeature in the hasher-API: Pass the digest in, so we don't have to touch the (potential) new objhead, unless we intend to insert it.   Poul-Henning Kamp
• [master] 4bbbc69 In theory diag_bitmap could change between (non-)assignment to t0 and use of t0. Always initialize t0 to something.   Poul-Henning Kamp
• [master] dbad89e Now that saint-mode uses the digest, we can remove the temporary objcore assignment.   Poul-Henning Kamp
• [master] ca3099c Lessen the load on oh->mtx a bit.   Poul-Henning Kamp
• [master] bbc0bc4 Start preparing hash-lookup for streaming objects   Poul-Henning Kamp
• [master] d734511 Update comment to match reality   Poul-Henning Kamp
• [master] 20d561d Redo the oc->NOTYET stuff after pondering.   Poul-Henning Kamp
• [master] bd0246a Edge closer to the new worldorder   Poul-Henning Kamp
• [master] c40a7d0 Unbusy the objectcore when we start fetching the body.   Poul-Henning Kamp
• [master] ed645d2 Remove busyobj from objcore under oh-mtx   Poul-Henning Kamp
• [master] 4edc10c The busyobj refcount must be protected by the oh->mtx just like the oc->busyobj pointer.   Poul-Henning Kamp
• [master] 4d33fab Introduce a couple of local busyobj variables for clarity   Poul-Henning Kamp
• [master] ab58597 Add a separate table for the http VSL headers   Poul-Henning Kamp
• [master] e4996ef Link the HTTP VSLs to the index in the http structure   Poul-Henning Kamp
• [master] 61367de Simplify the CPP magic involved in HTTP headers, now that we can rely on everybody agreeing about the order of these fields.   Poul-Henning Kamp
• [master] f177a35 Rework the way we log HTTP headers: Rather than the Rx/Tx/Obj, tag them with Req/Resp/Bereq/Beresp/Obj instead.   Poul-Henning Kamp
• [master] 745c828 Add two fields to the VSL tag definition, for documentation use.   Poul-Henning Kamp
• [master] 9fc9e7d Add short descriptions   Andreas Plesner Jacobsen
• [master] e37585d Make sure we distribute vsl_tags_http.h too   Tollef Fog Heen
• [master] 5a1b770 Adjust varnishncsa and varnishreplay for the new tag names   Tollef Fog Heen
• [master] fc8678c Remove unused tag   Andreas Plesner Jacobsen
• [master] c5a1c75 With streaming, we cannot allow a storage chunk to move once we have started to fill it. Add a param to the STV_trim() method to indicate acceptability of move.   Poul-Henning Kamp
• [master] 3a4edf1 Centralize the updating of the fetch_obj->len field.   Poul-Henning Kamp
• [master] 70468df add a label for refs   Per Buer
• [master] 56cfc90 alternative, simpler cookie stripper by James Light   Per Buer
• [master] f93c6fa kill warnings. fix a ref   Per Buer

Last message date: Thu Mar 29 09:46:09 UTC 2012
Archived on: Wed Feb 11 18:06:24 UTC 2026

• Messages sorted by: [ thread ] [ subject ] [ author ]
• More info on this list...