Fix mixed output of adapter and regular traces

[lslusarczyk]

Fix #2002 issue.
When merged this PR will make regular UR calls tracing being printed calls in two separate lines like PI does.

[pbalcer]

I'm not really a fan of this approach, I'd prefer to have indents like ltrace. But I guess it solves the immediate problem...

Also, please try to keep the first line of the commit message under 50 characters, and the rest of the message lines under ~70.

$ man git commit

...
Though not required, it’s a good idea to begin the commit message with a single short (less than 50 character)
line summarizing the change, followed by a blank line and then a more thorough description. The text up to the
first blank line in a commit message is treated as the commit title, and that title is used throughout Git.

[pbalcer]

not all the arguments can be accessed here, because some of them may be output arguments, with uninitialized content, which results in UB on access.

We either have to selectively access the parameters, adding an option to printFunctionParams to decide whether we are pre or post function execution, or, we can simply read those arguments after the execution. Which is what the original implementation did, for this exact reason.

[lslusarczyk]

@pbalcer , can you take a look again to this PR and merge this if all is OK?
There are three checks failing here:

1. Build and test / E2E L0 / Start e2e job / Build SYCL, UR, run E2E (L0, level_zero...) this is expected as llvm has no changes from [SYCL][E2E] fixed UR FileCheck conditions to match new traces intel/llvm#15439 yet
2. Build and test / E2E CUDA / Start e2e job / Build SYCL, UR, run E2E (CUDA...) expected - same situation
3. Build and test / E2E OpenCL / Start e2e job / Build SYCL, UR, run E2E (OPENCL...) this fail is unrelated to my change and as I see this check is executed very rarely and even if it runs then nobody cares about failures in it and happily merges PRs with red color on it

• see checks in this merged PR: Add support for command-buffer kernel updates #1924 (most recent one which had to run this check, August)
• or this one Add two new properties to ur_kernel_group_info_t #1996

E2E tests on llvm pass on my change and corresponding change in llvm is approved, see: intel/llvm#15439