Checkout Tools
  • last updated 2 hours ago
Constraints: committers
Constraints: files
Constraints: dates
Remove mod_tls. All of the filters in mod_tls can now be found in

mod_ssl. Also, mod_ssl is being kept up to date, while mod_tls is not,

so that if you use mod_tls, there are a lot of bugs, both security related,

and otherwise.

Toggle the /Zi flag to allow all supportd versions of VC (5, 6, 7) to

build 'out of the box' in debugging mode.

  1. … 31 more files in changeset.

Changes for rbb's #include/build openssl/inc32 patch.

  1. … 2 more files in changeset.

Based on the VC6 switch - and pulling the win32 make files from cvs

- some more ignores are in order.

  1. … 32 more files in changeset.

The purpose of this patch is to toggle the debugging mode (default) to

Program Database (from Program Database for Modify on the fly debugging).

The net effect of this patch is to clean up all of the irrelevant entries

associated with either the debugging or release command line switches, and

generally straighten the projects as they would be exported from VC6/SP5.

The outcome of this patch is that VC5 users -should- be able to load and

build the workspace without any errors (as they used to have no symbols

database at all, the /ZI option doesn't work, they had to use

to toggle these to /Zi.)

  1. … 28 more files in changeset.

Given a choice between chasing my tail to keep these updated, and rolling

a set when we actually roll a tarball, I'll take the later, thank you.

  1. … 28 more files in changeset.

Ignore goodness

  1. … 30 more files in changeset.

More Versioning

  1. … 16 more files in changeset.

Update for the 2.0.23 tag

  1. … 26 more files in changeset.

Undo my unintentionally overzelous name change, and fix the style of sizeof()

Port mod_tls.c from Eastern Europe to Apache-land

Ooops... my fooness.

  1. … 1 more file in changeset.

Ok... for _today_ ...

This patch conditionally builds against openssl 0.9.6b _if_ openssl is

unpacked and properly built (using pretty much the defaults) in the

srclib/openssl/ directory. Someday soon this needs to be more exhaustive,

but this should solve the 80% problem :)

  1. … 2 more files in changeset.

Change lib locations for openssl's debugging default targets

  1. … 1 more file in changeset.
zero the SSLStateMachine in case of early error, else SSLStateMachine_free segvs

Update for latest structure changes for 2.0.21

  1. … 21 more files in changeset.
s/destroy/free/g; likeness to SSL naming

give some more diagnostics if server cert or key file cannot be read

add cleanup of SSLStateMachine to tls filter

Fix incl/libpath to / slashes for an upcoming patch to cvtdsp.c

  1. … 1 more file in changeset.

Remove openssl path-specific .h comparisons

Refresh .mak files for tagging 2.0.19

  1. … 22 more files in changeset.

Ignore goodness

Just a mssing brace

tls goodness for Win32

Win32/OS2 require APR_STATUS_IS_EFOO() tests, not == EFOO, since our

socket error codes vary.

Be more strict when --with-tls was given without a DIR value.

Try to auto-detect the location of the OpenSSL stuff. The old logic was

incorrect anyway (it tested for .../ssl/ssl.h and then used .../include/ssl.h).

Now we have a configurable list of candidates, each with:

a name, an include path, a lib path, and a list of libs.

Back out the recent change to ap_get_brigade, to make it use indirection

again. The problem is that the amount of data read from the network,

is not necessarily the amount of data returned from the filters. It is

possible for input filters to add bytes to the data read from the network.

To fix the original bug, I just removed the line from ap_get_client_block

that decremented r->remaining, we allow the http_filter to do that for


I have also removed an incorrect comment.

  1. … 13 more files in changeset.
Fix a bug in the input handling. ap_http_filter() was modifying *readbytes

which corresponded to r->remaining (in ap_get_client_block). However,

ap_get_client_block was *also* adjusting r->remaining. Net result was that

PUT (and probably POST) was broken. (at least on large inputs)

To fix it, I simply removed the indirection on "readbytes" for input

filters. There is no reason for them to return data (the brigade length is

the return length). This also simplifies a number of calls where people

needed to do &zero just to pass zero.

I also added a number of comments about operations and where things could be

improved, or are (semi) broken.

  1. … 11 more files in changeset.