https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
Mark Thomas changed:
What|Removed |Added
Status|NEW |RESOLVED
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #9 from Mark Thomas ---
I've back-ported this to 8.5.x for 8.5.1 and 8.0.x for 8.0.34.
I'll hold off on back-porting this to 7.0.x until after the 7.0.69 tag.
--
You are receiving this mail because:
You are the
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #8 from Mark Thomas ---
Fixed in trunk for 9.0.0.M5 and the unit tests still all pass.
--
You are receiving this mail because:
You are the assignee for the bug.
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #7 from Mark Thomas ---
isAsyncStarted() makes sense and if we code to that and it is later relaxed we
won't need to change anything. I think I am most of the way to a working patch
for this.
We also need to
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #6 from Remy Maucherat ---
isAsyncStarted is correct, and it's not the same as asyncContext != null, I
checked GF before writing comment 3. However; Tomcat needs to know if a
startAsync has been called earlier for
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #5 from Mark Thomas ---
Reading the Javadoc, the test is 'has one of the startAsync() methods been
called' which is not quite the same as isAsyncStarted() == false. The spec
document isn't much better. It uses the
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #4 from Mark Thomas ---
On the plus side, the async code is reasonably well covered by the unit tests.
The down side, as Rémy points out is that we'll need to carefully review all
the calls to getAsyncContext().
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #3 from Remy Maucherat ---
Right now the spec is that getAsyncContext() should throw an ISE if
isAsyncStarted returns false (which is not the same as asyncContext != null).
This is an extremely risky change, some
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #2 from Remy Maucherat ---
Maybe that is what the specification says, but using null in this situation is
considerably better than using an ISE which should be reserved for some
invalid/meaningless situations.
Bad
https://bz.apache.org/bugzilla/show_bug.cgi?id=59261
--- Comment #1 from Rob Winch ---
I messed up the description some. Sorry about that. It should read:
If ServletRequest.getAsyncContext() is invoked and has not been put into
asynchronous mode an IllegalStateException should
10 matches
Mail list logo