Permalink
Switch branches/tags
Commits on Aug 30, 2011
  1. Bump version numbers for 3.0.1

    Tollef Fog Heen committed Aug 30, 2011
  2. Add 3.0.1 changes

    Tollef Fog Heen committed Aug 30, 2011
  3. Move varnishd parameter documentation into their own file

    Having all the parameters in their own file makes updating them
    simpler.
    Tollef Fog Heen committed Aug 30, 2011
  4. Add missing ABI line check that got lost

    Tollef Fog Heen committed Aug 30, 2011
  5. Make -x dumprst always available

    Tollef Fog Heen committed Aug 30, 2011
  6. Clean up VSM setup if mmap fails

    Fixes #995 (the rest of it).
    Rogier R. Mulhuijzen committed with Tollef Fog Heen Aug 30, 2011
  7. Reintroduce ExpKill VSL records.

    Fixes		#946
    
    Patch by	scoof
    bsdphk committed with Tollef Fog Heen Aug 30, 2011
  8. Don't look at a static version of the VSM during startup.

    Fixes		#995
    Patch by	DocWilco
    bsdphk committed with Tollef Fog Heen Aug 30, 2011
  9. Document vcl_init{} and vcl_fini{}

    Fixes	#990
    bsdphk committed with Tollef Fog Heen Aug 30, 2011
  10. Don't allow "error" in vcl_init{} and vcl_fini{}, they have no

    request to error.
    
    Fixes	#996
    bsdphk committed with Tollef Fog Heen Aug 30, 2011
  11. Fix time_firstbyte bug in varnishncsa

    Jean-Baptiste Quenot committed with Tollef Fog Heen Aug 29, 2011
  12. Add rudimentary -X support to varnishncsa

    Avoid crashing if an -X parameter is passed, and skip any requests
    where the URL or Host matches the -X parameter.
    Tollef Fog Heen committed Aug 30, 2011
  13. Administrative fixes:

    - Cleanup whitespace bogons
    - Make all machine-generated no-copyright notices consistent
    - Update copyrights
    bsdphk committed with Tollef Fog Heen Aug 29, 2011
Commits on Aug 29, 2011
  1. Use SMA instead in the hope that we can escape what I think is a

    page-size issue on ppc64
    bsdphk committed with Tollef Fog Heen Aug 29, 2011
  2. Reduce http_req_size on 32 bit machines

    Fixes	#988
    bsdphk committed with Tollef Fog Heen Aug 26, 2011
  3. Recent changes to the RFC2616 ttl calculation confused what ttl

    value was actually being used.
    
    Fixed by:	DocWilco
    Fixes		#984
    bsdphk committed with Tollef Fog Heen Aug 26, 2011
  4. Documentation fixes for 3.0

    scoof committed with Tollef Fog Heen Aug 25, 2011
  5. Add note about thread_pool_max

    scoof committed with Tollef Fog Heen Aug 16, 2011
  6. Add a "fallback" director as a variant of round-robin.

    This one always picks the first healty backend, in the order
    they are specified in the VCL.
    
    Submitted by:	DocWilco
    bsdphk committed with Tollef Fog Heen Aug 25, 2011
  7. Add missing header for string concatention operator

    cosimo committed with Tollef Fog Heen Aug 25, 2011
  8. Be much more bombastic about the per-request flags.

    This highlights that they really need to go into a struct or
    bitmap for clarity, but I'm not doing that right before 3.0.1
    
    Fixes	#986
    
    Many Thanks To:	Kristian
    bsdphk committed with Tollef Fog Heen Aug 24, 2011
  9. Add various flags to the panic output

    bsdphk committed with Tollef Fog Heen Aug 24, 2011
  10. Even harder asserts for kristians test

    bsdphk committed with Tollef Fog Heen Aug 24, 2011
  11. Move some asserts up in the code, to catch Kristians panic.

    bsdphk committed with Tollef Fog Heen Aug 24, 2011
Commits on Aug 25, 2011
  1. Update configure etc for rc1

    Tollef Fog Heen committed Aug 25, 2011
Commits on Aug 24, 2011
  1. Test stevedore/LRU/Nuke when the body allocation fails

    bsdphk committed with Tollef Fog Heen Aug 22, 2011
Commits on Aug 23, 2011
  1. SMA->trim() did not return the trimmed space to the pool causing

    the SMA to run out of space eventually.
    bsdphk committed with Tollef Fog Heen Aug 22, 2011
  2. Document 3.0 changes

    Tollef Fog Heen committed Aug 19, 2011
  3. Lock panic buffer before mangling it

    If we were panic-ing in multiple threads at the same time, the panic
    buffer would be partially overwritten.  Prevent this with a mutex
    Tollef Fog Heen committed Aug 19, 2011
  4. An assert to make sure...

    bsdphk committed with Tollef Fog Heen Aug 22, 2011
  5. Adjust c0004[45] limits to work on 32 bit

    Tollef Fog Heen committed Aug 23, 2011
  6. Allow larger bodies in varnishtest

    bsdphk committed with Tollef Fog Heen Aug 22, 2011
  7. The law of unconsidered consequences strikes again:

    When we pushed the object allocation into the stevedores for
    -spersistent, we did not add LRU eviction to that allocation
    path.
    
    Then we added the Transient storage as a fallback for objects
    we could not allocate, and they all went there.
    
    Change the way object allocation works as follows:
    
            If VCL set a stevedore hint and it is valid, we stick
            with it, and LRU that stevedore attempting to make space.
    
            If no valid hint is given, try all stevedores in turn,
            then LRU one of them to make space.
    
    Fixes	#953
    bsdphk committed with Tollef Fog Heen Aug 22, 2011
  8. Properly clean up if we bail out of cnt_error because we cannot allocate

    an object.
    
    Fixes	#985
    bsdphk committed with Tollef Fog Heen Aug 22, 2011