The only h2 psuedo header that wasn't previously possible to change by a
user. This change also makes it impossible to send a HTTP/1 header that
starts with a colon, which I don't think anyone does anyway.
The other pseudo headers are possible to change indirectly by doing the
rightly crafted request.
Reported-by: siddharthchhabrap on github
Fixes#8381Closes#8393
There has been no TPF related changes done since September 2010 (commit
7e1a45e224) and since this is a platform that is relatively different
than many others (== needs attention), I draw the conclusion that this
build is broken since a long time.
Closes#8378
Fix a bug that does not require a new CVE as discussed on hackerone.com.
Previously `connection_id` was accessed after returning connection to
the shared pool.
Bug: https://hackerone.com/reports/1463013Closes#8355
The callback gets called because of the call to PR_Recv() done to
attempt to avoid RST on the TCP connection. The conn->bundle pointer is
already cleared at this point so avoid dereferencing it.
Reported-by: Eric Musser
Fixes#8341Closes#8342
mbedtls/certs.h file contains only certificates example (all definitions
is beginning by mbedtls_test_*). None of them is used so we can avoid
include the file.
Closes#8343
The TLS backends convert the host name to SNI name and need to use that.
This involves cutting off any trailing dot and lowercasing.
Co-authored-by: Jay Satiro
Closes#8320
Reverts 5de8d84098 (May 2014, shipped in 7.37.0) and the
follow-up changes done afterward.
Keep the dot in names for everything except the SNI to make curl behave
more similar to current browsers. This means 'name' and 'name.' send the
same SNI for different 'Host:' headers.
Updated test 1322 accordingly
Fixes#8290
Reported-by: Charles Cazabon
Closes#8320
Avoids the need to clone the strings before check, thus avoiding
mallocs, which for cases where there are many SAN names in a cert could
end up numerous.
Closes#8321
- Wrap OpenSSL / wolfSSL MD5 functions instead of taking their function
addresses during static initialization.
Depending on how curl was built the old way may have used a dllimport
function address during static initialization, which is not standard
compliant, resulting in Visual Studio warning C4232 (nonstandard
extension). Instead the function pointers now point to the wrappers
which call the MD5 functions.
This change only affects OpenSSL and wolfSSL because calls to other SSL
libraries' md5 functions were already wrapped. Also sha256.c already
does this for all SSL libraries.
Ref: https://github.com/curl/curl/pull/8298
Closes https://github.com/curl/curl/pull/8318
The tools.ietf.org domain has been deprecated a while now, with the
links being redirected to datatracker.ietf.org.
Rather than make people eat that redirect time, this change switches the
URL to a more canonical source.
Closes#8317
put all #include of openssl files behind wolfssl ifdefs so that we can
use the wolfssl/ prefixed include paths. Without these curl only builds
when wolfssl is built with enable-all.
Fixes#8292Closes#8315
quiche has just switched it's qlog serialization format to JSON-SEQ by
default . The spec says this SHOULD use `.sqlog` extension.
I believe ngtcp2 also supports JSON-SEQ by default as of
9baf06fc3f
Let's update curl so that tools know what format we are using!
Closes#8316
Previously curl would just silently ignore it if the necessary defines
are not present at build-time.
Reported-by: Stefan Eissing
Fixes#8309Closes#8310
1. The function would only ever return CURLE_OK anyway
2. Only one caller actually used the return code
3. Most callers did (void)Curl_disconnect()
Closes#8303
This makes most libcurl functions return error if called from within a
callback using the same multi handle. For example timer or socket
callbacks calling curl_multi_socket_action.
Reported-by: updatede on github
Fixes#8282Closes#8286
Follow-up to bbf8cae44d
We removed support for the watcom builds files back in September
2020. This removes all remaining watcom references and ifdefs.
Closes#8287
