[llvm] [objdump] Fix dxcontainer tests on big endian host (PR #163427)

[Chris B via llvm-commits]

llvm-beanz wrote:

> Looks like this might have broken little endian systems...

This wasn't related to endian-ness, it is that some implementations of `od` in base systems have different numbers of leading 0's on the address.
 
> The premerge checks failed on the PR before landing as well. Are you able to quickly fix forward or revert?

Fix is in.

The pre-merge checks here are a bit tricky to grok. AArch64 succeeded, but the Linux x86 failed and the failure message is buried in the middle of an extremely long build log. I _did_ look at the Linux failure before I hit the merge button, but the log ended with no clear indication of which sub-action failed just an opaque python script reported failure. That led me to (incorrectly) assume this was a tooling error not caused by my change.

It would be nice if we could split the build tasks so that the logs get broken up into smaller pieces where we can reasonably expect that the failures would be at the end instead of the middle.



https://github.com/llvm/llvm-project/pull/163427