History log of /trafficserver/iocore/eventsystem/P_IOBuffer.h (Results 1 - 25 of 109)
Revision (<<< Hide revision tags) (Show revision tags >>>) Date Author Comments
# 72299397 14-May-2020 Leif Hedstrom

Update existingh formatting with new clang-format package


Revision tags: 8.0.7, 7.1.10, 7.1.9, 7.1.9-rc1, 8.0.6, 7.1.9-rc0, 8.0.6-rc0
# 4edbda64 07-Nov-2019 Masaori Koshiba

Remove unused functions of IOBuffer


# 64595af3 02-Sep-2019 Masaori Koshiba

Cleanup: unifdef TRACK_BUFFER_USER


# c40271ae 03-Sep-2019 Masaori Koshiba

Reduce unnecesary IOBufferBlock allocation

- Add unit tests for MIOBuffer::write_avail()
- Do nothing when the next block of the current writer exists


# c65f0a1f 29-Aug-2019 Masaori Koshiba

Cleanup: Remove unused AllocType and unused functions


Revision tags: 8.0.5, 7.1.8, 8.0.4, 7.1.7, 7.1.7-rc0, 8.0.4-rc0, 8.0.3, 7.1.6, 8.0.2, 8.0.1, 7.1.5, 7.1.5-rc0, 8.0.1-rc0, 8.0.0, 8.0.0-rc3, 8.0.0-rc2, 8.0.0-rc1, 8.0.0-rc0
# affba2de 07-Aug-2018 Randall Meyer

Fixes spelling in iocore


# 7651e269 19-Mar-2019 Masaori Koshiba

Ran clang-tidy with modernize-use-default-member-init

1. Remove unmatched #endif and set `_location` nullptr.
- iocore/eventsystem/I_IOBuffer.h
- iocore/eventsystem/P_IOBuffer.h

Ran clang-tidy with modernize-use-default-member-init

1. Remove unmatched #endif and set `_location` nullptr.
- iocore/eventsystem/I_IOBuffer.h
- iocore/eventsystem/P_IOBuffer.h

2. Include "ts/apidefs.h"
- include/tscpp/api/Stat.h

3. Rollback changes made with -fix-error option (clang-diagnostic-implicit-function-declaration)
- plugins/lua/ts_lua_client_request.c
- plugins/lua/ts_lua_client_response.c
- plugins/lua/ts_lua_server_response.c
- plugins/experimental/remap_stats/remap_stats.c

show more ...


# 841697c6 23-Jan-2019 Alan M. Carroll

Add IOBufferReader::block_read_view method.


# 079a4027 30-Aug-2018 Susan Hinrichs

PR #3724: Restructure library headers to unify include paths.


Revision tags: 7.1.4, 6.2.3, 7.1.4-rc0
# 7621f3ed 15-May-2018 Alan M. Carroll

IOBufferChain: A class for holding an IO buffer block chain.


# c29e2916 25-Apr-2018 Leif Hedstrom

Updated to new version of clang-format


# c9ca6476 24-Apr-2018 Bryan Call

Ran clang-tidy with readability-braces-around-statements


# 35f4804e 12-Apr-2018 Bryan Call

Ran clang-tidy with modernize-loop-convert


# b5037ade 11-Apr-2018 Bryan Call

Ran clang-tidy with modernize-use-nullptr


# b82d0aaf 11-Apr-2018 Masakazu Kitajo

Use #pragma once


Revision tags: 7.1.3
# c7b2c645 13-Jan-2018 Gancho Tenev

Reset water_mark in new_empty_MIOBuffer_internal

Moving water_mark reset one level up from MIOBuffer::alloc()
to new_MIOBuffer_internal() and adding it "symmetrically" to
new_empty_M

Reset water_mark in new_empty_MIOBuffer_internal

Moving water_mark reset one level up from MIOBuffer::alloc()
to new_MIOBuffer_internal() and adding it "symmetrically" to
new_empty_MIOBuffer_internal().
This will allow new_empty_MIOBuffer() to benefit from the
fix in commit db81103d0 as well.

show more ...


# db81103d 12-Jan-2018 Gancho Tenev

new_MIOBuffer: uninitialized water_mark fix.

While calling new_MIOBuffer() the water_mark is not initialized,
water_mark could end up with a very high (random) value to be used
durin

new_MIOBuffer: uninitialized water_mark fix.

While calling new_MIOBuffer() the water_mark is not initialized,
water_mark could end up with a very high (random) value to be used
during the network reads. This always results in overly aggressive
uninterrupted reading from the connection, severe buffer size increase
and CPU saturation of one or more ET_NET threads which renders the
traffic server unresponsive. Initializing the water_mark fixes the issue.

The problem can be easily reproduced by a series of very large TLS POST
requests in combination with a stalling origin server and periodic use
of FetchSM which internally allocates a buffer using new_MIOBuffer()
and sets water_mark to INT64_MAX.
While serving the POST requests a buffer previously used by FetchSM
would be reused and its water_mark would end up being equal to IN64_MAX
since the water mark is not initialized when calling new_MIOBuffer().
This always results in aggressive uninterrupted network reads, the
buffer size keep growing progressively and the constant block iteration
in IOBufferReader::read_avail() at some point would saturate the ET_NET
thread CPU.

This patch should fix issue #2144

Additionally fixed erroneous #ifdef TS_USE_PORT directive in
SSLNetVConnection::net_read_io().

show more ...


Revision tags: 7.1.2, 7.1.2-rc3, 7.1.2-rc2, 7.1.2-rc1, 7.1.2-rc0, 7.1.1, 7.1.1-rc0, 6.2.2, 6.2.2-rc0, 7.1.0, 7.1.0-rc0
# 6bdc94e2 15-Feb-2017 Leif Hedstrom

Cleans up ICC options, and some build issues

This removes a number of the previous warning exclusions that we
we used to do. We then also have to clean up the core code in a
few plac

Cleans up ICC options, and some build issues

This removes a number of the previous warning exclusions that we
we used to do. We then also have to clean up the core code in a
few places, sometimes looking slightly odd (e.g. the meaningless
const that we have in a few places).

show more ...


Revision tags: 6.2.1
# 2c732409 13-Dec-2016 Susan Hinrichs

TS-5095: IOBufferReader::read_avail adds to CPU utilization


Revision tags: 7.0.0, 7.0.0-rc2
# 2ae0497d 26-Oct-2016 Nathan Garabedian

TS-5009 CID 1022011 Logically dead code removal


# 662be1f8 07-Dec-2016 Oknet Xu

TS-5082: define IOBufferReader::is_read_avail_more_than with TS_INLINE keyword


Revision tags: 7.0.0-rc1, 7.0.0-rc0
# ced4da13 14-Oct-2016 James Peach

TS-4977: Prefer nullptr to NULL.


Revision tags: 6.2.0, 6.2.0-rc2, 6.2.0-rc1
# d6906e2a 19-Jun-2016 Leif Hedstrom

TS-4565 Update indentation with = alignment


Revision tags: 6.2.0-rc0
# cf53450b 06-May-2016 James Peach

TS-4425: Switch IOBuffers over to Ptr::get().


# c32379b6 06-May-2016 Phil Sorber

TS-4418: Update code with new .clang-format config change


12345