large request body result in wasm crash

[wbpcode]

If you are reporting any crash or any potential security issue, do not
open an issue in this repo. Please report the issue via emailing
envoy-security@googlegroups.com where the issue will be triaged appropriately.

Title: One line description

Description:

#36411 fixed a bug where the wasm filter may cannot get complete request body. But it also introduce another new bug. Given a big request, the addDecodedData() in the decode/encodeData() may trigger the buffer high watermark callback and result in local reply sending (413 payload too large). And after the local reply is sent, the decode/encodeData() will continue run and call into the vm. But because the local reply is sent so the context in vm may has been removed, the calling into vm may finally result in a crash/panic (depends on the sdk implementation).