Multiprocessors and Multithreading Write a sequence of MIPS assembly instruction

Question

Multiprocessors and Multithreading Write a sequence of MIPS assembly instructions using LL/SC pairs to correctly and concurrently insert a node (immediately after the current node) into a shared singly linked list without using locks. You may assume that the only operations permitted on the shared list are traversal, read-node-value, add-node, and modify-node-value (note the lack of a delete operation, which simplifies the above task considerably). Consider the code sequences used in the consistency supplement to show that TSO can produce different outcomes than sequential consistency. Construct a new set of code sequences utilizing the same 3 binary output variables for which TSO and sequential consistency still differ, but not all 8 output values are valid under TSO. Insert the minimum number of mfence operations into your code in 10b in order to make TSO behave equivalently to sequential consistency. One of the newest modifications to the x 86 ISA is the introduction of transactional memory primitives. Discuss how the memory transactions supported by these new instructions relate to atomic primitives, consistency models and coherence.

Explanation / Answer

#include #include /* Function to assign `Naive Bayes'-style weights to each element of each document vector. */ void bow_naivebayes_set_weights (bow_barrel *barrel) { int ci; bow_cdoc *cdoc; int wi; /* a "word index" into WI2DVF */ int max_wi; /* the highest "word index" in WI2DVF. */ bow_dv *dv; /* the "document vector" at index WI */ int dvi; /* an index into the DV */ int weight_setting_num_words = 0; /* We assume that we have already called BOW_BARREL_NEW_VPC() on BARREL, so BARREL already has one-document-per-class. */ assert (!strcmp (barrel->method->name, "naivebayes") || !strcmp (barrel->method->name, "crossentropy")); max_wi = MIN (barrel->wi2dvf->size, bow_num_words()); /* The CDOC->PRIOR should have been set in bow_barrel_new_vpc(); verify it. */ for (ci = 0; ci cdocs->length; ci++) { cdoc = bow_array_entry_at_index (barrel->cdocs, ci); assert (cdoc->prior >= 0); } #if 0 /* For Shumeet, make all counts either 0 or 1. */ for (wi = 0; wi wi2dvf, wi); if (dv == NULL) continue; for (dvi = 0; dvi length; dvi++) { assert (dv->entry[dvi].count); dv->entry[dvi].count = 1; } } /* And set uniform priors */ for (ci = 0; ci cdocs->length; ci++) { cdoc = bow_array_entry_at_index (barrel->cdocs, ci); cdoc->prior = 1.0; } #endif /* Get the total number of terms in each class; store this in CDOC->WORD_COUNT. */ for (ci = 0; ci cdocs->length; ci++) { cdoc = bow_array_entry_at_index (barrel->cdocs, ci); cdoc->word_count = 0; } for (wi = 0; wi wi2dvf, wi); if (dv == NULL) continue; for (dvi = 0; dvi length; dvi++) { cdoc = bow_array_entry_at_index (barrel->cdocs, dv->entry[dvi].di); cdoc->word_count += dv->entry[dvi].count; } } /* Set the weights in the BARREL's WI2DVF so that they are equal to P(w|C), the probability of a word given a class. */ for (wi = 0; wi wi2dvf, wi); /* If the model doesn't know about this word, skip it. */ if (dv == NULL) continue; /* Now loop through all the elements, setting their weights */ for (dvi = 0; dvi length; dvi++) { cdoc = bow_array_entry_at_index (barrel->cdocs, dv->entry[dvi].di); /* Here CDOC->WORD_COUNT is the total number of words in the class */ /* We use Laplace Estimation. */ dv->entry[dvi].weight = ((float) (1 + dv->entry[dvi].count) / (barrel->wi2dvf->num_words + cdoc->word_count)); assert (dv->entry[dvi].weight > 0); } weight_setting_num_words++; /* Set the IDF. NaiveBayes doesn't use it; make it have no effect */ dv->idf = 1.0; } #if 0 fprintf (stderr, "wi2dvf num_words %d, weight-setting num_words %d ", barrel->wi2dvf->num_words, weight_setting_num_words); #endif } /* For changing weight of unseen words. I really should implement `deleted interpolation' */ /* M_EST_P summed over all words in the vocabulary must sum to 1.0! */ #if 0 /* This is the special case of the M-estimate that is `Laplace smoothing' */ #define M_EST_M (barrel->wi2dvf->num_words) #define M_EST_P (1.0 / barrel->wi2dvf->num_words) #define WORD_PRIOR_COUNT 1.0 #else #define M_EST_M (cdoc->word_count ? (((float)barrel->wi2dvf->num_words) / cdoc->word_count) : 1.0) #define M_EST_P (1.0 / barrel->wi2dvf->num_words) #endif int bow_naivebayes_score_loo (bow_barrel *barrel, bow_wv *query_wv, bow_score *bscores, int bscores_len, int loo_class) { double *scores; /* will become prob(class), indexed over CI */ int ci; /* a "class index" (document index) */ int wvi; /* an index into the entries of QUERY_WV. */ int dvi; /* an index into a "document vector" */ float pr_w_c; /* P(w|C), prob a word is in a class */ double pr_tf; /* P(w|C)^TF, ditto, by occurr's in QUERY_WV */ double log_pr_tf; /* log(P(w|C)^TF), ditto, log() of it */ double rescaler; /* Rescale SCORES by this after each word */ double new_score; /* a temporary holder */ int num_scores; /* number of entries placed in SCORES */ /* Allocate space to store scores for *all* classes (documents) */ scores = alloca (barrel->cdocs->length * sizeof (double)); /* Instead of multiplying probabilities, we will sum up log-probabilities, (so we don't loose floating point resolution), and then take the exponent of them to get probabilities back. */ /* Initialize the SCORES to the class prior probabilities. */ if (bow_print_word_scores) printf ("%s ", "(CLASS PRIOR PROBABILIES)"); for (ci = 0; ci cdocs->length; ci++) { bow_cdoc *cdoc; cdoc = bow_array_entry_at_index (barrel->cdocs, ci); if (bow_uniform_class_priors) scores[ci] = 1; else { /* LOO_CLASS is not implemented for cases in which we are not doing uniform class priors. */ assert (loo_class == -1); assert (cdoc->prior > 0.0f && cdoc->prior prior); if (((bow_params_naivebayes*)(barrel->method->params)) ->score_with_log_probabilities == bow_yes) scores[ci] = - scores[ci]; } assert (scores[ci] > -FLT_MAX + 1.0e5); if (bow_print_word_scores) printf ("%16s %-40s %10.9f ", "", (strrchr (cdoc->filename, '/') ? : cdoc->filename), scores[ci]); } /* Loop over each word in the word vector QUERY_WV, putting its contribution into SCORES. */ for (wvi = 0; wvi num_entries; wvi++) { int wi; /* the word index for the word at WVI */ bow_dv *dv; /* the "document vector" for the word WI */ /* Get information about this word. */ wi = query_wv->entry[wvi].wi; dv = bow_wi2dvf_dv (barrel->wi2dvf, wi); /* If the model doesn't know about this word, skip it. */ if (!dv) continue; if (bow_print_word_scores) printf ("%-30s (queryweight=%.8f) ", bow_int2word (wi), query_wv->entry[wvi].weight * query_wv->normalizer); rescaler = DBL_MAX; /* Loop over all classes, putting this word's (WI's) contribution into SCORES. */ for (ci = 0, dvi = 0; ci cdocs->length; ci++) { /* Both these values are pretty arbitrary small numbers. */ static const double min_pr_tf = FLT_MIN * 1.0e5; bow_cdoc *cdoc; cdoc = bow_array_entry_at_index (barrel->cdocs, ci); assert (cdoc->type == model); /* Assign PR_W_C to P(w|C), either using a DV entry, or, if there is no DV entry for this class, using M-estimate smoothing */ if (dv) while (dvi length && dv->entry[dvi].di entry[dvi].di == ci) { if (loo_class == ci) { /* xxx This is not exactly right, because BARREL->WI2DVF->NUM_WORDS might have changed with the removal of QUERY_WV's document. */ pr_w_c = ((float) ((M_EST_M * M_EST_P) + dv->entry[dvi].count - query_wv->entry[wvi].count) / (M_EST_M + cdoc->word_count - query_wv->entry[wvi].count)); assert (pr_w_c > 0 && pr_w_c entry[dvi].count) / (M_EST_M + cdoc->word_count)); assert (pr_w_c > 0 && pr_w_c WI2DVF->NUM_WORDS might have changed with the removal of QUERY_WV's document. */ pr_w_c = ((M_EST_M * M_EST_P) / (M_EST_M + cdoc->word_count - query_wv->entry[wvi].count)); assert (pr_w_c > 0 && pr_w_c word_count)); assert (pr_w_c > 0 && pr_w_c 0 && pr_w_c entry[wvi].count); /* PR_TF can be zero due to round-off error, when PR_W_C is very small and QUERY_WV->ENTRY[CURRENT_INDEX].COUNT is very large. Here we fudgingly avoid this by insisting that PR_TF not go below some arbitrary small number. */ if (pr_tf -FLT_MAX + 1.0e5); if (((bow_params_naivebayes*)(barrel->method->params)) ->score_with_log_probabilities == bow_yes) log_pr_tf = -log_pr_tf; scores[ci] += log_pr_tf; if (bow_print_word_scores) printf (" %8.2e %7.2f %-40s %10.9f ", pr_w_c, log_pr_tf, (strrchr (cdoc->filename, '/') ? : cdoc->filename), scores[ci]); /* Keep track of the minimum score updated for this word. */ if (rescaler > scores[ci]) rescaler = scores[ci]; } if (((bow_params_naivebayes*)(barrel->method->params)) ->score_with_log_probabilities == bow_no) { /* Loop over all classes, re-scaling SCORES so that they don't get so small we loose floating point resolution. This scaling always keeps all SCORES positive. */ if (rescaler < 0) { for (ci = 0; ci cdocs->length; ci++) { /* Add to SCORES to bring them close to zero. RESCALER is expected to often be less than zero here. */ /* xxx If this doesn't work, we could keep track of the min and the max, and sum by their average. */ scores[ci] += -rescaler; assert (scores[ci] > -DBL_MAX + 1.0e5 && scores[ci] method->params)) ->score_with_log_probabilities == bow_no) { /* Rescale the SCORE one last time, this time making them all 0 or negative, so that exp() will work well, especially around the higher-probability classes. */ { rescaler = -DBL_MAX; for (ci = 0; ci cdocs->length; ci++) if (scores[ci] > rescaler) rescaler = scores[ci]; /* RESCALER is now the maximum of the SCORES. */ for (ci = 0; ci cdocs->length; ci++) scores[ci] -= rescaler; } /* Use exp() on the SCORES to get probabilities from log-probabilities. */ for (ci = 0; ci cdocs->length; ci++) { new_score = exp (scores[ci]); /* assert (new_score > 0 && new_score length; ci++) scores[ci] = 1.0 / scores[ci]; } /* Normalize the SCORES so they all sum to one. */ { double scores_sum = 0; for (ci = 0; ci cdocs->length; ci++) scores_sum += scores[ci]; for (ci = 0; ci cdocs->length; ci++) { scores[ci] /= scores_sum; /* assert (scores[ci] > 0); */ } } /* Return the SCORES by putting them (and the `class indices') into SCORES in sorted order. */ { num_scores = 0; for (ci = 0; ci cdocs->length; ci++) { if (num_scores

Related Questions

Navigate

• Browse (All)
• Browse M
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.