diff --git a/include/jemalloc/internal/arena_inlines_b.h b/include/jemalloc/internal/arena_inlines_b.h
index ea69a688..96889c1b 100644
--- a/include/jemalloc/internal/arena_inlines_b.h
+++ b/include/jemalloc/internal/arena_inlines_b.h
@@ -15,7 +15,8 @@ arena_t *arena_aalloc(tsdn_t *tsdn, const void *ptr);
 size_t arena_salloc(tsdn_t *tsdn, const void *ptr);
 size_t arena_vsalloc(tsdn_t *tsdn, const void *ptr);
 void arena_dalloc_no_tcache(tsdn_t *tsdn, void *ptr);
-void arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool slow_path);
+void arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache,
+    dalloc_ctx_t *dalloc_ctx, bool slow_path);
 void arena_sdalloc_no_tcache(tsdn_t *tsdn, void *ptr, size_t size);
 void arena_sdalloc(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache,
     bool slow_path);
@@ -194,7 +195,8 @@ arena_dalloc_no_tcache(tsdn_t *tsdn, void *ptr) {
 }
 
 JEMALLOC_ALWAYS_INLINE void
-arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool slow_path) {
+arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache,
+    dalloc_ctx_t *dalloc_ctx, bool slow_path) {
 	assert(!tsdn_null(tsdn) || tcache == NULL);
 	assert(ptr != NULL);
 
@@ -203,13 +205,21 @@ arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool slow_path) {
 		return;
 	}
 
-	rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsdn_tsd(tsdn));
 	szind_t szind;
 	bool slab;
-	rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx, (uintptr_t)ptr,
-	    true, &szind, &slab);
+	rtree_ctx_t *rtree_ctx;
+	if (dalloc_ctx != NULL) {
+		szind = dalloc_ctx->szind;
+		slab = dalloc_ctx->slab;
+		assert(szind != NSIZES);
+	} else {
+		rtree_ctx = tsd_rtree_ctx(tsdn_tsd(tsdn));
+		rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx,
+		    (uintptr_t)ptr, true, &szind, &slab);
+	}
 
 	if (config_debug) {
+		rtree_ctx = tsd_rtree_ctx(tsdn_tsd(tsdn));
 		extent_t *extent = rtree_extent_read(tsdn, &extents_rtree,
 		    rtree_ctx, (uintptr_t)ptr, true);
 		assert(szind == extent_szind_get(extent));
diff --git a/include/jemalloc/internal/arena_structs_b.h b/include/jemalloc/internal/arena_structs_b.h
index 00e0d0c8..b6ba3dbc 100644
--- a/include/jemalloc/internal/arena_structs_b.h
+++ b/include/jemalloc/internal/arena_structs_b.h
@@ -256,4 +256,10 @@ struct arena_tdata_s {
 	ticker_t		decay_ticker;
 };
 
+/* Used to pass rtree lookup context down the deallocation path. */
+struct dalloc_ctx_s {
+	szind_t szind;
+	bool slab;
+};
+
 #endif /* JEMALLOC_INTERNAL_ARENA_STRUCTS_B_H */
diff --git a/include/jemalloc/internal/arena_types.h b/include/jemalloc/internal/arena_types.h
index ba53c408..435b930d 100644
--- a/include/jemalloc/internal/arena_types.h
+++ b/include/jemalloc/internal/arena_types.h
@@ -19,6 +19,7 @@ typedef struct arena_decay_s arena_decay_t;
 typedef struct arena_bin_s arena_bin_t;
 typedef struct arena_s arena_t;
 typedef struct arena_tdata_s arena_tdata_t;
+typedef struct dalloc_ctx_s dalloc_ctx_t;
 
 typedef enum {
 	percpu_arena_disabled = 0,
diff --git a/include/jemalloc/internal/jemalloc_internal.h.in b/include/jemalloc/internal/jemalloc_internal.h.in
index 62dae0c4..17c77aa3 100644
--- a/include/jemalloc/internal/jemalloc_internal.h.in
+++ b/include/jemalloc/internal/jemalloc_internal.h.in
@@ -1095,8 +1095,8 @@ void *ipalloct(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero,
     tcache_t *tcache, arena_t *arena);
 void *ipalloc(tsd_t *tsd, size_t usize, size_t alignment, bool zero);
 size_t ivsalloc(tsdn_t *tsdn, const void *ptr);
-void idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool is_internal,
-    bool slow_path);
+void idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache,
+    dalloc_ctx_t *dalloc_ctx, bool is_internal, bool slow_path);
 void idalloc(tsd_t *tsd, void *ptr);
 void isdalloct(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache,
     bool slow_path);
@@ -1188,8 +1188,8 @@ ivsalloc(tsdn_t *tsdn, const void *ptr) {
 }
 
 JEMALLOC_ALWAYS_INLINE void
-idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool is_internal,
-    bool slow_path) {
+idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache, dalloc_ctx_t *dalloc_ctx,
+    bool is_internal, bool slow_path) {
 	assert(ptr != NULL);
 	assert(!is_internal || tcache == NULL);
 	assert(!is_internal || arena_ind_get(iaalloc(tsdn, ptr)) <
@@ -1201,12 +1201,12 @@ idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool is_internal,
 	if (!is_internal && *tsd_reentrancy_levelp_get(tsdn_tsd(tsdn)) != 0) {
 		tcache = NULL;
 	}
-	arena_dalloc(tsdn, ptr, tcache, slow_path);
+	arena_dalloc(tsdn, ptr, tcache, dalloc_ctx, slow_path);
 }
 
 JEMALLOC_ALWAYS_INLINE void
 idalloc(tsd_t *tsd, void *ptr) {
-	idalloctm(tsd_tsdn(tsd), ptr, tcache_get(tsd), false, true);
+	idalloctm(tsd_tsdn(tsd), ptr, tcache_get(tsd), NULL, false, true);
 }
 
 JEMALLOC_ALWAYS_INLINE void
diff --git a/src/ckh.c b/src/ckh.c
index 463f8dd1..03262ef5 100644
--- a/src/ckh.c
+++ b/src/ckh.c
@@ -282,12 +282,12 @@ ckh_grow(tsd_t *tsd, ckh_t *ckh) {
 		ckh->lg_curbuckets = lg_curcells - LG_CKH_BUCKET_CELLS;
 
 		if (!ckh_rebuild(ckh, tab)) {
-			idalloctm(tsd_tsdn(tsd), tab, NULL, true, true);
+			idalloctm(tsd_tsdn(tsd), tab, NULL, NULL, true, true);
 			break;
 		}
 
 		/* Rebuilding failed, so back out partially rebuilt table. */
-		idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, NULL, true, true);
 		ckh->tab = tab;
 		ckh->lg_curbuckets = lg_prevbuckets;
 	}
@@ -329,7 +329,7 @@ ckh_shrink(tsd_t *tsd, ckh_t *ckh) {
 	ckh->lg_curbuckets = lg_curcells - LG_CKH_BUCKET_CELLS;
 
 	if (!ckh_rebuild(ckh, tab)) {
-		idalloctm(tsd_tsdn(tsd), tab, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), tab, NULL, NULL, true, true);
 #ifdef CKH_COUNT
 		ckh->nshrinks++;
 #endif
@@ -337,7 +337,7 @@ ckh_shrink(tsd_t *tsd, ckh_t *ckh) {
 	}
 
 	/* Rebuilding failed, so back out partially rebuilt table. */
-	idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, true, true);
+	idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, NULL, true, true);
 	ckh->tab = tab;
 	ckh->lg_curbuckets = lg_prevbuckets;
 #ifdef CKH_COUNT
@@ -418,7 +418,7 @@ ckh_delete(tsd_t *tsd, ckh_t *ckh) {
 	    (unsigned long long)ckh->nrelocs);
 #endif
 
-	idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, true, true);
+	idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, NULL, true, true);
 	if (config_debug) {
 		memset(ckh, JEMALLOC_FREE_JUNK, sizeof(ckh_t));
 	}
diff --git a/src/jemalloc.c b/src/jemalloc.c
index 513d9cd2..bc659325 100644
--- a/src/jemalloc.c
+++ b/src/jemalloc.c
@@ -313,7 +313,7 @@ a0ialloc(size_t size, bool zero, bool is_internal) {
 
 static void
 a0idalloc(void *ptr, bool is_internal) {
-	idalloctm(TSDN_NULL, ptr, false, is_internal, true);
+	idalloctm(TSDN_NULL, ptr, NULL, NULL, is_internal, true);
 }
 
 void *
@@ -2045,21 +2045,29 @@ ifree(tsd_t *tsd, void *ptr, tcache_t *tcache, bool slow_path) {
 	assert(ptr != NULL);
 	assert(malloc_initialized() || IS_INITIALIZER);
 
+	dalloc_ctx_t dalloc_ctx;
+	rtree_ctx_t *rtree_ctx = tsd_rtree_ctx(tsd);
+	rtree_szind_slab_read(tsd_tsdn(tsd), &extents_rtree, rtree_ctx,
+	    (uintptr_t)ptr, true, &dalloc_ctx.szind, &dalloc_ctx.slab);
+	assert(dalloc_ctx.szind != NSIZES);
+
 	size_t usize;
 	if (config_prof && opt_prof) {
-		usize = isalloc(tsd_tsdn(tsd), ptr);
+		usize = index2size(dalloc_ctx.szind);
 		prof_free(tsd, ptr, usize);
 	} else if (config_stats) {
-		usize = isalloc(tsd_tsdn(tsd), ptr);
+		usize = index2size(dalloc_ctx.szind);
 	}
 	if (config_stats) {
 		*tsd_thread_deallocatedp_get(tsd) += usize;
 	}
 
 	if (likely(!slow_path)) {
-		idalloctm(tsd_tsdn(tsd), ptr, tcache, false, false);
+		idalloctm(tsd_tsdn(tsd), ptr, tcache, &dalloc_ctx, false,
+		    false);
 	} else {
-		idalloctm(tsd_tsdn(tsd), ptr, tcache, false, true);
+		idalloctm(tsd_tsdn(tsd), ptr, tcache, &dalloc_ctx, false,
+		    true);
 	}
 }
 
diff --git a/src/prof.c b/src/prof.c
index db1ef035..40610d71 100644
--- a/src/prof.c
+++ b/src/prof.c
@@ -590,7 +590,7 @@ prof_gctx_try_destroy(tsd_t *tsd, prof_tdata_t *tdata_self, prof_gctx_t *gctx,
 		prof_leave(tsd, tdata_self);
 		/* Destroy gctx. */
 		malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock);
-		idalloctm(tsd_tsdn(tsd), gctx, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), gctx, NULL, NULL, true, true);
 	} else {
 		/*
 		 * Compensate for increment in prof_tctx_destroy() or
@@ -704,7 +704,7 @@ prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx) {
 	}
 
 	if (destroy_tctx) {
-		idalloctm(tsd_tsdn(tsd), tctx, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), tctx, NULL, NULL, true, true);
 	}
 }
 
@@ -736,8 +736,8 @@ prof_lookup_global(tsd_t *tsd, prof_bt_t *bt, prof_tdata_t *tdata,
 			if (ckh_insert(tsd, &bt2gctx, btkey.v, gctx.v)) {
 				/* OOM. */
 				prof_leave(tsd, tdata);
-				idalloctm(tsd_tsdn(tsd), gctx.v, NULL, true,
-				    true);
+				idalloctm(tsd_tsdn(tsd), gctx.v, NULL, NULL,
+				    true, true);
 				return true;
 			}
 			new_gctx = true;
@@ -761,7 +761,8 @@ prof_lookup_global(tsd_t *tsd, prof_bt_t *bt, prof_tdata_t *tdata,
 
 		if (tgctx.v != NULL) {
 			/* Lost race to insert. */
-			idalloctm(tsd_tsdn(tsd), tgctx.v, NULL, true, true);
+			idalloctm(tsd_tsdn(tsd), tgctx.v, NULL, NULL, true,
+			    true);
 		}
 	}
 	prof_leave(tsd, tdata);
@@ -833,7 +834,7 @@ prof_lookup(tsd_t *tsd, prof_bt_t *bt) {
 			if (new_gctx) {
 				prof_gctx_try_destroy(tsd, tdata, gctx, tdata);
 			}
-			idalloctm(tsd_tsdn(tsd), ret.v, NULL, true, true);
+			idalloctm(tsd_tsdn(tsd), ret.v, NULL, NULL, true, true);
 			return NULL;
 		}
 		malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock);
@@ -1245,7 +1246,7 @@ prof_gctx_finish(tsd_t *tsd, prof_gctx_tree_t *gctxs) {
 					tctx_tree_remove(&gctx->tctxs,
 					    to_destroy);
 					idalloctm(tsd_tsdn(tsd), to_destroy,
-					    NULL, true, true);
+					    NULL, NULL, true, true);
 				} else {
 					next = NULL;
 				}
@@ -1915,7 +1916,7 @@ prof_tdata_init_impl(tsd_t *tsd, uint64_t thr_uid, uint64_t thr_discrim,
 
 	if (ckh_new(tsd, &tdata->bt2tctx, PROF_CKH_MINITEMS, prof_bt_hash,
 	    prof_bt_keycomp)) {
-		idalloctm(tsd_tsdn(tsd), tdata, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), tdata, NULL, NULL, true, true);
 		return NULL;
 	}
 
@@ -1971,10 +1972,11 @@ prof_tdata_destroy_locked(tsd_t *tsd, prof_tdata_t *tdata,
 	assert(prof_tdata_should_destroy_unlocked(tdata, even_if_attached));
 
 	if (tdata->thread_name != NULL) {
-		idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, NULL, true,
+		    true);
 	}
 	ckh_delete(tsd, &tdata->bt2tctx);
-	idalloctm(tsd_tsdn(tsd), tdata, NULL, true, true);
+	idalloctm(tsd_tsdn(tsd), tdata, NULL, NULL, true, true);
 }
 
 static void
@@ -2171,7 +2173,8 @@ prof_thread_name_set(tsd_t *tsd, const char *thread_name) {
 	}
 
 	if (tdata->thread_name != NULL) {
-		idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, NULL, true,
+		    true);
 		tdata->thread_name = NULL;
 	}
 	if (strlen(s) > 0) {
diff --git a/src/tcache.c b/src/tcache.c
index 34b46afd..09776e1a 100644
--- a/src/tcache.c
+++ b/src/tcache.c
@@ -499,10 +499,10 @@ tcache_destroy(tsd_t *tsd, tcache_t *tcache, bool tsd_tcache) {
 		void *avail_array =
 		    (void *)((uintptr_t)tcache_small_bin_get(tcache, 0)->avail -
 		    (uintptr_t)tcache_bin_info[0].ncached_max * sizeof(void *));
-		idalloctm(tsd_tsdn(tsd), avail_array, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), avail_array, NULL, NULL, true, true);
 	} else {
 		/* Release both the tcache struct and avail array. */
-		idalloctm(tsd_tsdn(tsd), tcache, NULL, true, true);
+		idalloctm(tsd_tsdn(tsd), tcache, NULL, NULL, true, true);
 	}
 }