| Message ID | 20221019221321.3033920-2-coltonlewis@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | KVM: selftests: randomize memory access of dirty_log_perf_test | expand |
On Wed, Oct 19, 2022, Colton Lewis wrote: > Implement random number generation for guest code to randomize parts > of the test, making it less predictable and a more accurate reflection > of reality. > > Create a -r argument to specify a random seed. If no argument is > provided, the seed defaults to 0. The random seed is set with > perf_test_set_random_seed() and must be set before guest_code runs to > apply. > > The random number generator chosen is the Park-Miller Linear > Congruential Generator, a fancy name for a basic and well-understood > random number generator entirely sufficient for this purpose. Each > vCPU calculates its own seed by adding its index to the seed provided. > > Signed-off-by: Colton Lewis <coltonlewis@google.com> > Reviewed-by: Ricardo Koller <ricarkol@google.com> > Reviewed-by: David Matlack <dmatlack@google.com> This patch has changed a fair bit since David and Ricardo gave their reviews, their Reviewed-by's should be dropped. Alternatively, if a patch has is on the fence so to speak, i.e. has changed a little bit but not thaaaaat much, you can add something in the cover letter, e.g. "David/Ricardo, I kept your reviews, let me know if that's not ok". But in this case, I think the code has changed enough that their reviews should be dropped. > --- > .../testing/selftests/kvm/dirty_log_perf_test.c | 12 ++++++++++-- > .../selftests/kvm/include/perf_test_util.h | 2 ++ > tools/testing/selftests/kvm/include/test_util.h | 7 +++++++ > .../testing/selftests/kvm/lib/perf_test_util.c | 7 +++++++ > tools/testing/selftests/kvm/lib/test_util.c | 17 +++++++++++++++++ > 5 files changed, 43 insertions(+), 2 deletions(-) I think it makes sense to introduce "struct guest_random_state" separately from the usage in perf_test_util and dirty_log_perf_test. E.g. so that if we need to revert the perf_test_util changes (extremely unlikely), we can do so without having to wipe out the pRNG at the same time. Or so that someone can pull in the pRNG to their series without having to take a dependency on the other changes. > diff --git a/tools/testing/selftests/kvm/include/test_util.h b/tools/testing/selftests/kvm/include/test_util.h > index befc754ce9b3..9e4f36a1a8b0 100644 > --- a/tools/testing/selftests/kvm/include/test_util.h > +++ b/tools/testing/selftests/kvm/include/test_util.h > @@ -152,4 +152,11 @@ static inline void *align_ptr_up(void *x, size_t size) > return (void *)align_up((unsigned long)x, size); > } > > +struct guest_random_state { > + uint32_t seed; > +}; > + > +struct guest_random_state new_guest_random_state(uint32_t seed); > +uint32_t guest_random_u32(struct guest_random_state *state); > + > #endif /* SELFTEST_KVM_TEST_UTIL_H */ > diff --git a/tools/testing/selftests/kvm/lib/perf_test_util.c b/tools/testing/selftests/kvm/lib/perf_test_util.c > index 9618b37c66f7..5f0eebb626b5 100644 > --- a/tools/testing/selftests/kvm/lib/perf_test_util.c > +++ b/tools/testing/selftests/kvm/lib/perf_test_util.c > @@ -49,6 +49,7 @@ void perf_test_guest_code(uint32_t vcpu_idx) > uint64_t gva; > uint64_t pages; > int i; > + struct guest_random_state rand_state = new_guest_random_state(pta->random_seed + vcpu_idx); lib/perf_test_util.c: In function ‘perf_test_guest_code’: lib/perf_test_util.c:52:35: error: unused variable ‘rand_state’ [-Werror=unused-variable] 52 | struct guest_random_state rand_state = new_guest_random_state(pta->random_seed + vcpu_idx); | ^~~~~~~~~~ This belongs in the next path. I'd also prefer to split the declaration from the initialization as this is an unnecessarily long line, e.g. struct perf_test_args *pta = &perf_test_args; struct perf_test_vcpu_args *vcpu_args = &pta->vcpu_args[vcpu_idx]; struct guest_random_state rand_state; uint64_t gva; uint64_t pages; uint64_t addr; uint64_t page; int i; rand_state = new_guest_random_state(pta->random_seed + vcpu_idx);
On Thu, Oct 27, 2022, Colton Lewis wrote: > Sean Christopherson <seanjc@google.com> writes: > > > On Wed, Oct 19, 2022, Colton Lewis wrote: > > > Signed-off-by: Colton Lewis <coltonlewis@google.com> > > > Reviewed-by: Ricardo Koller <ricarkol@google.com> > > > Reviewed-by: David Matlack <dmatlack@google.com> > > > This patch has changed a fair bit since David and Ricardo gave their > > reviews, their Reviewed-by's should be dropped. Alternatively, if a patch > > has is on the fence so to speak, i.e. has changed a little bit but not > > thaaaaat much, you can add something in the cover letter, e.g. > > "David/Ricardo, I kept your reviews, let me know if that's not ok". But in > > this case, I think the code has changed enough that their reviews should be > > dropped. > > > I talked to Ricardo privately and he thought it was ok to leave the > names in this borderline case. Heh, damned if you do, damned if you don't. On a more serious note, this is why I avoid doing off-list reviews except for the most basic sanity checks. Ask two people for their opinion and inevitably you'll get two different answers :-) By asking and responding on-list, there's is (a) a paper trail and (b) a chance for others to object before the decision is "finalized". > IMO, changing this interface doesn't change anything important of what the > patch is doing. Right, but the code is different. E.g. hypothetically, if you botched something while reworking the code to fit the new interace, then it looks as if multiple people gave a thumbs up to broken code, which can cause a maintainer to not give the patch as much scrutiny as that might to a patch without reviews. A recent example that's somewhat similar is commit 4293ddb788c1 ("KVM: x86/mmu: Remove redundant spte present check in mmu_set_spte"), where a conflict during a rebase of a relatively simple patch was mishandled and resulted in a buggy commit with three Reviewed-by tags where none of the reviews were given for the buggy code. There's no guarantee the bug would have been caught if the Reviewed-by tags had been dropped, but a "hey, I dropped your reviews from patch XYZ" likely would have drawn eyeballs to the patch in question. > Nevertheless, I'll drop the names and ask them to reconfirm. FWIW, if you've confirmed offline that someone is ok keeping _their_ review, that's totally fine, though throwing a comment in the cover letter is probably a good idea in that case. That's a decent rule of thumb in general; if a decision was made off-list, make a note of it on-list to keep others in the loop. > > I think it makes sense to introduce "struct guest_random_state" separately > > from the usage in perf_test_util and dirty_log_perf_test. E.g. so that if > > we need to revert the perf_test_util changes (extremely unlikely), we can > > do so without having to wipe out the pRNG at the same time. Or so that > > someone can pull in the pRNG to their series without having to take a > > dependency on the other changes. > > Will do. Was attempting to avoid adding unused code in its own commit > according to > https://www.kernel.org/doc/html/latest/process/5.Posting.html > > "Whenever possible, a patch which adds new code should make that code > active immediately." Yeah, this is another "rule" that is sometimes subjective, e.g. in this case it conflicts with the "one change per patch" rule. Since the RNG code can't be made completely active without pulling in yet more code (the guest_random_u32() usage in the next path), the intended benefit of the "use immediately" rule isn't really achieved, and so forcing the issue is a net negative. > > > diff --git a/tools/testing/selftests/kvm/lib/perf_test_util.c > > > b/tools/testing/selftests/kvm/lib/perf_test_util.c > > > index 9618b37c66f7..5f0eebb626b5 100644 > > > --- a/tools/testing/selftests/kvm/lib/perf_test_util.c > > > +++ b/tools/testing/selftests/kvm/lib/perf_test_util.c > > > @@ -49,6 +49,7 @@ void perf_test_guest_code(uint32_t vcpu_idx) > > > uint64_t gva; > > > uint64_t pages; > > > int i; > > > + struct guest_random_state rand_state = > > > new_guest_random_state(pta->random_seed + vcpu_idx); > > > lib/perf_test_util.c: In function ‘perf_test_guest_code’: > > lib/perf_test_util.c:52:35: error: unused variable ‘rand_state’ > > [-Werror=unused-variable] > > 52 | struct guest_random_state rand_state = > > new_guest_random_state(pta->random_seed + vcpu_idx); > > | ^~~~~~~~~~ > > > This belongs in the next path. I'd also prefer to split the declaration > > from the > > initialization as this is an unnecessarily long line, e.g. > > Understood that this is implied by the previous comment. As for the line > length, checkpatch doesn't complain. Checkpatch now only complains if the line length is >100, but there is still an 80 column "soft" limit that is preferred by most of the kernel, KVM included. Checkpatch was relaxed because too many people were wrapping code in really weird places "because checkpatch complained", but preferred style is still to say under the soft limit. In other words, stay under the soft limit unless the alternative sucks worse :-)
diff --git a/tools/testing/selftests/kvm/dirty_log_perf_test.c b/tools/testing/selftests/kvm/dirty_log_perf_test.c index f99e39a672d3..c97a5e455699 100644 --- a/tools/testing/selftests/kvm/dirty_log_perf_test.c +++ b/tools/testing/selftests/kvm/dirty_log_perf_test.c @@ -132,6 +132,7 @@ struct test_params { bool partition_vcpu_memory_access; enum vm_mem_backing_src_type backing_src; int slots; + uint32_t random_seed; }; static void toggle_dirty_logging(struct kvm_vm *vm, int slots, bool enable) @@ -225,6 +226,9 @@ static void run_test(enum vm_guest_mode mode, void *arg) p->slots, p->backing_src, p->partition_vcpu_memory_access); + /* If no argument provided, random seed will be 1. */ + pr_info("Random seed: %u\n", p->random_seed); + perf_test_set_random_seed(vm, p->random_seed ? p->random_seed : 1); perf_test_set_wr_fract(vm, p->wr_fract); guest_num_pages = (nr_vcpus * guest_percpu_mem_size) >> vm->page_shift; @@ -352,7 +356,7 @@ static void help(char *name) { puts(""); printf("usage: %s [-h] [-i iterations] [-p offset] [-g] " - "[-m mode] [-n] [-b vcpu bytes] [-v vcpus] [-o] [-s mem type]" + "[-m mode] [-n] [-b vcpu bytes] [-v vcpus] [-o] [-r random seed ] [-s mem type]" "[-x memslots]\n", name); puts(""); printf(" -i: specify iteration counts (default: %"PRIu64")\n", @@ -380,6 +384,7 @@ static void help(char *name) printf(" -v: specify the number of vCPUs to run.\n"); printf(" -o: Overlap guest memory accesses instead of partitioning\n" " them into a separate region of memory for each vCPU.\n"); + printf(" -r: specify the starting random seed.\n"); backing_src_help("-s"); printf(" -x: Split the memory region into this number of memslots.\n" " (default: 1)\n"); @@ -406,7 +411,7 @@ int main(int argc, char *argv[]) guest_modes_append_default(); - while ((opt = getopt(argc, argv, "eghi:p:m:nb:f:v:os:x:")) != -1) { + while ((opt = getopt(argc, argv, "eghi:p:m:nb:f:v:or:s:x:")) != -1) { switch (opt) { case 'e': /* 'e' is for evil. */ @@ -442,6 +447,9 @@ int main(int argc, char *argv[]) case 'o': p.partition_vcpu_memory_access = false; break; + case 'r': + p.random_seed = atoi(optarg); + break; case 's': p.backing_src = parse_backing_src_type(optarg); break; diff --git a/tools/testing/selftests/kvm/include/perf_test_util.h b/tools/testing/selftests/kvm/include/perf_test_util.h index eaa88df0555a..f1050fd42d10 100644 --- a/tools/testing/selftests/kvm/include/perf_test_util.h +++ b/tools/testing/selftests/kvm/include/perf_test_util.h @@ -35,6 +35,7 @@ struct perf_test_args { uint64_t gpa; uint64_t size; uint64_t guest_page_size; + uint32_t random_seed; int wr_fract; /* Run vCPUs in L2 instead of L1, if the architecture supports it. */ @@ -52,6 +53,7 @@ struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int nr_vcpus, void perf_test_destroy_vm(struct kvm_vm *vm); void perf_test_set_wr_fract(struct kvm_vm *vm, int wr_fract); +void perf_test_set_random_seed(struct kvm_vm *vm, uint32_t random_seed); void perf_test_start_vcpu_threads(int vcpus, void (*vcpu_fn)(struct perf_test_vcpu_args *)); void perf_test_join_vcpu_threads(int vcpus); diff --git a/tools/testing/selftests/kvm/include/test_util.h b/tools/testing/selftests/kvm/include/test_util.h index befc754ce9b3..9e4f36a1a8b0 100644 --- a/tools/testing/selftests/kvm/include/test_util.h +++ b/tools/testing/selftests/kvm/include/test_util.h @@ -152,4 +152,11 @@ static inline void *align_ptr_up(void *x, size_t size) return (void *)align_up((unsigned long)x, size); } +struct guest_random_state { + uint32_t seed; +}; + +struct guest_random_state new_guest_random_state(uint32_t seed); +uint32_t guest_random_u32(struct guest_random_state *state); + #endif /* SELFTEST_KVM_TEST_UTIL_H */ diff --git a/tools/testing/selftests/kvm/lib/perf_test_util.c b/tools/testing/selftests/kvm/lib/perf_test_util.c index 9618b37c66f7..5f0eebb626b5 100644 --- a/tools/testing/selftests/kvm/lib/perf_test_util.c +++ b/tools/testing/selftests/kvm/lib/perf_test_util.c @@ -49,6 +49,7 @@ void perf_test_guest_code(uint32_t vcpu_idx) uint64_t gva; uint64_t pages; int i; + struct guest_random_state rand_state = new_guest_random_state(pta->random_seed + vcpu_idx); gva = vcpu_args->gva; pages = vcpu_args->pages; @@ -229,6 +230,12 @@ void perf_test_set_wr_fract(struct kvm_vm *vm, int wr_fract) sync_global_to_guest(vm, perf_test_args); } +void perf_test_set_random_seed(struct kvm_vm *vm, uint32_t random_seed) +{ + perf_test_args.random_seed = random_seed; + sync_global_to_guest(vm, perf_test_args.random_seed); +} + uint64_t __weak perf_test_nested_pages(int nr_vcpus) { return 0; diff --git a/tools/testing/selftests/kvm/lib/test_util.c b/tools/testing/selftests/kvm/lib/test_util.c index 6d23878bbfe1..c4d2749fb2c3 100644 --- a/tools/testing/selftests/kvm/lib/test_util.c +++ b/tools/testing/selftests/kvm/lib/test_util.c @@ -17,6 +17,23 @@ #include "test_util.h" +/* + * Random number generator that is usable from guest code. This is the + * Park-Miller LCG using standard constants. + */ + +struct guest_random_state new_guest_random_state(uint32_t seed) +{ + struct guest_random_state s = {.seed = seed}; + return s; +} + +uint32_t guest_random_u32(struct guest_random_state *state) +{ + state->seed = (uint64_t)state->seed * 48271 % ((uint32_t)(1 << 31) - 1); + return state->seed; +} + /* * Parses "[0-9]+[kmgt]?". */