| Message ID | 20240510224002.2324578-2-andrew.cooper3@citrix.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | x86/xen-cpuid: Use automagically generated names | expand |
On Fri, May 10, 2024 at 11:39:59PM +0100, Andrew Cooper wrote: > Rename INIT_FEATURE_NAMES to INIT_FEATURE_NAME_TO_VAL as we're about to gain a > inverse mapping of the same thing. > > Use dict.items() unconditionally. iteritems() is a marginal perf optimsiation > for Python2 only, and simply not worth the effort on a script this small. My understanding is that what used to be iteritems() in Python 2 is the behavior of items() in Python 3 (return a generator instead of a copy of the dictionary list). > Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com> Thanks, Roger.
On 14/05/2024 8:14 am, Roger Pau Monné wrote: > On Fri, May 10, 2024 at 11:39:59PM +0100, Andrew Cooper wrote: >> Rename INIT_FEATURE_NAMES to INIT_FEATURE_NAME_TO_VAL as we're about to gain a >> inverse mapping of the same thing. >> >> Use dict.items() unconditionally. iteritems() is a marginal perf optimsiation >> for Python2 only, and simply not worth the effort on a script this small. > My understanding is that what used to be iteritems() in Python 2 is > the behavior of items() in Python 3 (return a generator instead of a > copy of the dictionary list). Yes-ish. They're actually now view() objects following the official stabilisation of the internal format, which are more-efficient-still in the common case. > >> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> > Reviewed-by: Roger Pau Monné <roger.pau@citrix.com> Thanks. ~Andrew
diff --git a/tools/libs/light/libxl_cpuid.c b/tools/libs/light/libxl_cpuid.c index ce4f3c7095ba..063fe86eb72f 100644 --- a/tools/libs/light/libxl_cpuid.c +++ b/tools/libs/light/libxl_cpuid.c @@ -296,7 +296,7 @@ int libxl_cpuid_parse_config(libxl_cpuid_policy_list *policy, const char* str) {NULL, 0, NA, CPUID_REG_INV, 0, 0} }; - static const struct feature_name features[] = INIT_FEATURE_NAMES; + static const struct feature_name features[] = INIT_FEATURE_NAME_TO_VAL; /* * NB: if we switch to using a cpu_policy derived object instead of a * libxl_cpuid_policy_list we could get rid of the featureset -> cpuid leaf diff --git a/xen/arch/x86/cpu-policy.c b/xen/arch/x86/cpu-policy.c index 99871b8e0e05..b96f4ee55cc4 100644 --- a/xen/arch/x86/cpu-policy.c +++ b/xen/arch/x86/cpu-policy.c @@ -43,7 +43,7 @@ static const uint32_t deep_features[] = INIT_DEEP_FEATURES; static const struct feature_name { const char *name; unsigned int bit; -} feature_names[] __initconstrel = INIT_FEATURE_NAMES; +} feature_names[] __initconstrel = INIT_FEATURE_NAME_TO_VAL; /* * Parse a list of cpuid feature names -> bool, calling the callback for any diff --git a/xen/tools/gen-cpuid.py b/xen/tools/gen-cpuid.py index 380b9d973a67..79d7f5c8e1c9 100755 --- a/xen/tools/gen-cpuid.py +++ b/xen/tools/gen-cpuid.py @@ -459,15 +459,10 @@ def write_results(state): state.output.write( """} -#define INIT_FEATURE_NAMES { \\ +#define INIT_FEATURE_NAME_TO_VAL { \\ """) - try: - _tmp = state.values.iteritems() - except AttributeError: - _tmp = state.values.items() - - for name, bit in sorted(_tmp): + for name, bit in sorted(state.values.items()): state.output.write( ' { "%s", %sU },\\\n' % (name, bit) )
Rename INIT_FEATURE_NAMES to INIT_FEATURE_NAME_TO_VAL as we're about to gain a inverse mapping of the same thing. Use dict.items() unconditionally. iteritems() is a marginal perf optimsiation for Python2 only, and simply not worth the effort on a script this small. Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> --- CC: Jan Beulich <JBeulich@suse.com> CC: Roger Pau Monné <roger.pau@citrix.com> v3: * New --- tools/libs/light/libxl_cpuid.c | 2 +- xen/arch/x86/cpu-policy.c | 2 +- xen/tools/gen-cpuid.py | 9 ++------- 3 files changed, 4 insertions(+), 9 deletions(-)