Add OR operator for option '--with-extra-multilib-test'

[Incarnation-p-lee]

This patch allows you provide the flags in build flags array acts on arch-abi respectively, you can use ',' to separate them. For example:

rv64gcv-lp64d:--param=riscv-autovec-lmul=dynamic,--param=riscv-autovec-preference=fixed-vlmax

will be consider as two target boards same as below:

riscv-sim/-march=rv64gcv/-mabi=lp64d/-mcmodel=medlow/--param=riscv-autovec-preference=fixed-vlmax
riscv-sim/-march=rv64gcv/-mabi=lp64d/-mcmodel=medlow/--param=riscv-autovec-lmul=dynamic

However, you can also leverage AND(:), OR(,) operator together but the OR(,) will always have the higher priority. For example:

rv64gcv-lp64d:--param=riscv-autovec-lmul=dynamic:--param=riscv-autovec-preference=fixed-vlmax,--param=riscv-autovec-lmul=m2

will be consider as tow target boars same as below:

riscv-sim/-march=rv64gcv/-mabi=lp64d/-mcmodel=medlow/--param=riscv-autovec-lmul=dynamic/--param=riscv-autovec-preference=fixed-vlmax
riscv-sim/-march=rv64gcv/-mabi=lp64d/-mcmodel=medlow/--param=riscv-autovec-lmul=m2

[cmuellner]

If I understood correctly, then this is not affecting the build, but only testing configurations.
If that's the case, then I wonder if we really need to make the interpretation of the --with-extra-multilib-test string more powerful, when we could simply add a mechanism that overloads the default behavior.

E.g. an environment variable EXTRA_MULTILIB_TEST could be used to override the current multilib test configuration. To test multiple such configurations, multiple invocations of make report with the appropriately set variable can be used. The calling context (e.g. a bash script, python script or a CI machinery) could then have all kinds of magic (of Turing complete languages) to generate the list and iterate over that.

I do not think this PR is targeting an invalid use case or is too complex, but I think it is not sufficient to express everything users may want to express to generate these multilib test lists. So we may end up with a PR introducing paratheses support in the future to overcome the limitations introduced by operator priorities...

[Incarnation-p-lee]

If I understood correctly, then this is not affecting the build, but only testing configurations. If that's the case, then I wonder if we really need to make the interpretation of the --with-extra-multilib-test string more powerful, when we could simply add a mechanism that overloads the default behavior.

E.g. an environment variable EXTRA_MULTILIB_TEST could be used to override the current multilib test configuration. To test multiple such configurations, multiple invocations of make report with the appropriately set variable can be used. The calling context (e.g. a bash script, python script or a CI machinery) could then have all kinds of magic (of Turing complete languages) to generate the list and iterate over that.

I do not think this PR is targeting an invalid use case or is too complex, but I think it is not sufficient to express everything users may want to express to generate these multilib test lists. So we may end up with a PR introducing paratheses support in the future to overcome the limitations introduced by operator priorities...

Thank you for the comments. Actually, this PR comes from one end-user requirement from #1368, and the AND / OR is somehow idempotent with limited effect to support.

I am totally OK if there is an overall solution for this.

[cmuellner]

You are right. Since this only affects the generate_target_board, the impact of adding new features is limited and merging this PR might be a reasonable choice.

@kito-cheng: any preferences?

[cmuellner]

I reopened this PR since I closed it by accident.

[Incarnation-p-lee]

Thanks @cmuellner and never mind. Let's wait for @kito-cheng 's comments on this.

[kito-cheng]

LGTM, seems like I missed this somehow

[Incarnation-p-lee]

Great, thanks @kito-cheng .