Re: [PATCH 04/12] sim/testsuite/cris/c: Use -sim3 but only for newlib targets

[Dimitar Dimitrov <dimitar@dinux.eu>]

On Tue, Feb 15, 2022 at 12:02:55AM +0100, Hans-Peter Nilsson via Gdb-patches wrote:
> Commit a39487c6685f "sim: cris: use -sim with C tests for cris-elf
> targets" caused " -sim" to be appended to CFLAGS_FOR_TARGET for
> cris*-*-elf, where testing had until then relied on
> "RUNTESTFLAGS=--target_board=cris-sim" being passed when running "make
> check-sim", adding the right options.  While "-sim" happens to work,
> the baseboard-file cris-sim.exp uses "-sim3" so for consistency use
> that instead.
> 
> Then commit b42f20d2ac72 "sim: testsuite: drop most specific istarget
> checks" caused " -sim" to be appended for *all* targets, which just
> doesn't work.  For example, for crisv32-linux-gnu, that's not a
> recognized option and will cause a dejagnu error and further testing
> in c.exp will be aborted.
> 
> While cris-sim.exp appends "-static" for *-linux-gnu, further changes
> in the test-suite have caused "linux"-specific tests to break, so that
> part will be tended to separately.
> 
> But, save and restore CFLAGS_FOR_TARGET around the modification and
> use where needed, to not have the CRIS-specific modification affect a
> continuing test-run (possibly for other targets).
> 
> sim/testsuite/cris:
> 	* c/c.exp (CFLAGS_FOR_TARGET): Replace appended option " -sim"
> 	with " -sim3", but do it conditionally for newlib targets.  Save
> 	and restore CFLAGS_FOR_TARGET in saved_CFLAGS_FOR_TARGET such
> 	that it doesn't affect the value of CFLAGS_FOR_TARGET outside
> 	c.exp.
> ---
>  sim/testsuite/cris/c/c.exp | 16 ++++++++++++----
>  1 file changed, 12 insertions(+), 4 deletions(-)
> 
> diff --git a/sim/testsuite/cris/c/c.exp b/sim/testsuite/cris/c/c.exp
> index 5711fd2f0bcd..3e186e072d8e 100644
> --- a/sim/testsuite/cris/c/c.exp
> +++ b/sim/testsuite/cris/c/c.exp
> @@ -17,6 +17,9 @@
>  
>  sim_init
>  
> +global global_cc_works
> +global global_cc_os

Hi,

FYI, this change introduces an error for "make check-sim" when configured
for pru-elf target. Error is gone if I revert this commit only. Note
that I don't have cross-CC in my PATH, which judging from sim_init_toolchain
function in sim-defs.exp is supported.

ERROR: -------------------------------------------
ERROR: in testcase /home/dinux/projects/pru/testbot-workspace/binutils/sim/testsuite/./cris/c/c.exp
ERROR:  can't read "global_cc_os": no such variable
ERROR:  tcl error code TCL LOOKUP VARNAME global_cc_os
ERROR:  tcl error info:
can't read "global_cc_os": no such variable
    while executing
"if { $global_cc_os == "newlib" } {
    append CFLAGS_FOR_TARGET " -sim3"
}"
    (file "/home/dinux/projects/pru/testbot-workspace/binutils/sim/testsuite/./cris/c/c.exp" line 35)


Regards,
Dimitar

> +
>  set CFLAGS_FOR_TARGET "-O2"
>  if [istarget cris-*-*] {
>      set mach "crisv10"
> @@ -24,13 +27,16 @@ if [istarget cris-*-*] {
>      set mach "crisv32"
>  }
>  
> -if [istarget *] {
> -    append CFLAGS_FOR_TARGET " -sim"
> +# Make sure we're using the right runtime for simulator runs.  If the
> +# cris-sim dejagnu baseboard is used, -sim3 will be duplicated, but
> +# that's ok.  For e.g. cris*-linux-gnu, neither -sim not -sim3 are
> +# supported options and likely not other targets too.
> +set saved_CFLAGS_FOR_TARGET $CFLAGS_FOR_TARGET
> +if { $global_cc_os == "newlib" } {
> +    append CFLAGS_FOR_TARGET " -sim3"
>  }
>  
>  # Using target_compile, since it is less noisy,
> -global global_cc_works
> -global global_cc_os
>  if { $global_cc_works == 1 } {
>      # Now check if we can link a program dynamically, and where
>      # libc.so is located.  If it is, we provide a sym link to the
> @@ -239,3 +245,5 @@ foreach src [lsort [glob -nocomplain $srcdir/$subdir/*.c]] {
>  	$status "$mach $testname"
>      }
>  }
> +
> +set CFLAGS_FOR_TARGET $saved_CFLAGS_FOR_TARGET
> -- 
> 2.30.2
>