> "Anthony" == Anthony Green <[EMAIL PROTECTED]> writes:
Anthony> This patch removes a bogus security check from ServerSocket.accept(),
Anthony> and replaces it with a request to implement a proper check. Once
Anthony> applied I will file a bug report for our records.
I filed this after we t
This patch removes a bogus security check from ServerSocket.accept(),
and replaces it with a request to implement a proper check. Once
applied I will file a bug report for our records.
This patch, along with all the other patches I've posted today, plus a
libgcj specific patch I posted, plus a nu
On Mon, 2005-12-26 at 19:07 -0700, Tom Tromey wrote:
> This looks ok to me.
>
> Anthony> +if (
> ((SocketChannel)key.channel()).isConnected ())
>
> Spacing here is wrong.
Thanks. Applied with the following corrected patch...
2005-12-26 Anthony Green <[EMAIL PROTECTE
On Mon, 2005-12-26 at 18:58 -0700, Tom Tromey wrote:
> > "Anthony" == Anthony Green <[EMAIL PROTECTED]> writes:
>
> Anthony> This next patch computes right amount of data to read from a
> Anthony> socket to stuff in a Buffer. When dst.limit < dst.capacity,
> Anthony> we're trying to put too m
> "Anthony" == Anthony Green <[EMAIL PROTECTED]> writes:
Anthony> After coming up with this, I discovered that this is
Anthony> basically the same solution that Stanley Brown came up with
Anthony> over a year
Anthony> ago... http://gcc.gnu.org/ml/java/2004-09/msg00130.html
I hate it when we
> "Anthony" == Anthony Green <[EMAIL PROTECTED]> writes:
Anthony> This first patch stops us from closing a socket when we get
Anthony> an exception during connect(). It is standard behaviour to
Anthony> get timeout exception here for non-blocking socket connect().
Anthony> We will never get a
> "Anthony" == Anthony Green <[EMAIL PROTECTED]> writes:
Anthony> This next patch computes right amount of data to read from a
Anthony> socket to stuff in a Buffer. When dst.limit < dst.capacity,
Anthony> we're trying to put too much data into the buffer and we get
Anthony> a buffer overflow
We weren't properly recording OP_CONNECT events in our select routine.
After coming up with this, I discovered that this is basically the same
solution that Stanley Brown came up with over a year ago...
http://gcc.gnu.org/ml/java/2004-09/msg00130.html
2005-12-26 Anthony Green <[EMAIL PROTECTE
This next patch computes right amount of data to read from a socket to
stuff in a Buffer. When dst.limit < dst.capacity, we're trying to put
too much data into the buffer and we get a buffer overflow exception.
Ok to apply?
Thanks,
AG
2005-12-26 Anthony Green <[EMAIL PROTECTED]>
*
I've been debugging various problems with Azureus. This is the first of
several patches.
This first patch stops us from closing a socket when we get an exception
during connect(). It is standard behaviour to get timeout exception
here for non-blocking socket connect(). We will never get a prop
I've attached my most recent version of KerberosPrincipal. I have
addressed most of the comments by Casey and Mark as well as comments
from others online.
There are still several FIXMEs in the code, but I don't think any of
them are barriers to adding the class to CVS. Most of them are details
t
11 matches
Mail list logo