This avoids a regression for issues like #334 where high speed
repeated connections eventually run the HTTP client out of
sockets because all of theirs end up in time_wait.
Maybe the trade-off here is suboptimal, but if both choices will
fail then we prefer fewer changes until the root cause is solved.
- now logs if -rootcertificates="" was used to disable payment request
authentication via X.509 certificates
- also logs which file is used as trusted root cert, if -rootcertificates
is set
1dd8ee7 improve tests for #5655 (Jonas Schnelli)
56c1093 fix tests for #5655 (Wladimir J. van der Laan)
16a5c18 Add a -rpckeepalive and disable RPC use of HTTP persistent connections. (Gregory Maxwell)
- this is based on #4122 (which can be closed)
Currently a payment request is only checked for expiration upon receipt.
It should be checked again immediately before sending coins to prevent
the user from paying to an expired invoice which would then require a
customer service interaction.
- add static verifyExpired() function to PaymentServer to be able to use
the same validation code in GUI and unit-testing code
- extend unit tests to use that function and also add an unit test which
overflows, because payment requests allow expires as uint64, whereas we
use int64_t for verification of expired payment requests
It turns out that some miners have been staying with old versions of
Bitcoin Core because their software behaves poorly with persistent
connections and the Bitcoin Core thread and connection limits.
What happens is that underlying HTTP libraries leave connections open
invisibly to their users and then the user runs into the default four
thread limit. This looks like Bitcoin Core is unresponsive to RPC.
There are many things that should be improved in Bitcoin Core's behavior
here, e.g. supporting more concurrent connections, not tying up threads
for idle connections, disconnecting kept-alive connections when limits
are reached, etc. All are fairly big, risky changes.
Disabling keep-alive is a simple workaround. It's often not easy to turn
off the keep-alive support in the client where it may be buried in some
platform library.
If you are one of the few who really needs persistent connections you
probably know that you want them and can find a switch; while if you
don't and the misbehavior is hitting you it is hard to discover the
source of your problems is keepalive related. Given that it is best
to default to off until they're handled better.
- Check that image contents match pre- and post- crushing.
- Also remove use of external tool to compute sha256 in favor of hashlib.
- contrib: remove all use of shell=True in strip_pngs.py
Using `shell=True` can be a security hazard. See e.g.
https://docs.python.org/2/library/subprocess.html#subprocess.check_output
17005bc [Qt] add payment request unit test for non matching networks (Philip Kaufmann)
080da96 [Qt] prepare paymentservertests for new unit tests (Philip Kaufmann)
- verify that payment request network matches client network
- add static verifyNetwork() function to PaymentServer to be able to use
the same validation code in GUI and unit-testing code
- add a second PaymentRequest Test CA certificate to paymentrequestdata.h
(serial number f0:da:97:e4:38:d7:64:16) as caCert2_BASE64
- rename existing Test CA certificate to caCert1_BASE64
- rename existing payment request data to know they belong to
caCert1_BASE64
- update comments to reflect the changes and add a missing comment to one
of the payment requests
b468e81 Qt: Clarify sign/verify dialog text to specifically state that these messages only prove one receives with the address in question, and makes no claim to sender of transactions (Luke Dashjr)
This harmonizes the block fetch timeout with the existing ping timeout
and eliminates a guaranteed eventual failure from congestion collapse
for a network operating right at its limit.
It's unlikely that we wouldn't suffer other failures if we were really
anywhere near the network's limit, and a complete avoidance of congestion
collapse risk requires (I think) an exponential back-off. So this isn't
a major concern, but I think it's also useful for reducing the complexity
of understanding out timeouts.
- it is helpful to be able to test and verify payment request processing
by allowing self signed root certificates (e.g. generated by Gavins
"certificate authority in a box")
- This option is just shown in the UI options, if -help-debug is enabled.
- before it was possible to use the steps to change e.g. amouns of
authenticated or unauthenticated payment requests (AmountSpinBox is
already set to read-only here) - this is now fixed
- also move the reimplemented stepEnabled() function to the
protected section of our class, where it belongs (see Qt doc)
8dccba6 fail immediately on an empty signature (Cory Fields)
dad7764 depends: bump openssl to 1.0.1k (Cory Fields)
488ed32 consensus: guard against openssl's new strict DER checks (Cory Fields)
New versions of OpenSSL will reject non-canonical DER signatures. However,
it'll happily decode them. Decode then re-encode before verification in order
to ensure that it is properly consumed.
Makes it possible to compactly provide a delibrately invalid signature
for use with CHECK(MULTI)SIG. For instance with BIP19 if m != n invalid
signatures need to be provided in the scriptSig; prior to this change
those invalid signatures would need to be large DER-encoded signatures.
Note that we may want to further expand on this change in the future by
saying that only OP_0 is a "valid" invalid signature; BIP19 even with
this change is inherently malleable as the invalid signatures can be any
validly encoded DER signature.
Gregory Maxwell