[Imprint][Data Privacy Statement][E-Mail]



View long theorem info. View the proof. Back to main index. View the specification. View theorem base summary, View theorem base listing, the used properties by spec, the dependency list, the symbol table,

Proof step 1 from proof rely-D2-D3 (conclusion) (back to tree)

⊦ ls.tid ≠ tid0 ∧ ( ( listp(gs.orl + gs.qrl, gs.vs, gs.ps, gs.owns) ∧ ¬ dups(gs.orl + gs.qrl) ∧ markedp(gs.orl, gs.ps) ∧ markedp(gs.orl, gs.vs) ∧ unmarkedp(gs.qrl, gs.ps) ∧ gs.orl ≠ [] ∧ gs.phead ∈ gs.orl ∧ rvownerok(gs.vretRefs, gs.vrvHp, gs.rvowns) ∧ allocrv(gs.vretRefs, gs.vrvHp) ∧ nonullRetRefs(gs.orl + gs.qrl, gs.vs, gs.vretRefs, gs.pretRefs) ∧ rvownerok(gs.pretRefs, gs.prvHp, gs.rvowns) ∧ allocrv(gs.pretRefs, gs.prvHp) ∧ dom gs.vrvHp = dom gs.prvHp ∧ ( gs.vs[gs.vhead].next ≠ null ∧ gs.vs[gs.vs[gs.vhead].next].deqID ≠ None → gs.rvowns(gs.vretRefs(gs.vs[gs.vs[gs.vhead].next].deqID.tid)) = Claimed(gs.vs[gs.vs[gs.vhead].next].deqID.tid, gs.vs[gs.vs[gs.vhead].next].val)) ∧ ( gs.orl.last = gs.vhead ∧ (gs.qrl ≠ [] → unmarkedp(gs.qrl.tail, gs.vs)) ∨ gs.orl.butlast ≠ [] ∧ gs.orl.butlast.last = gs.vhead ∧ gs.phead ≠ gs.orl.last ∧ unmarkedp(gs.qrl, gs.vs)) ∧ ( gs.vtail = (gs.orl + gs.qrl).last ∨ gs.qrl ≠ [] ∧ gs.vtail = (gs.orl + gs.qrl.butlast).last ∧ gs.vs[gs.qrl.last].deqID = None ∧ gs.vs[(gs.orl + gs.qrl).last].next = null)) ∧ nextnottid(ls.tid, gs.vhead, gs.vs) ∧ noid(ls.tid, gs.orl + gs.qrl, gs.vhead, gs.vs) ∧ noowns(ls.tid, gs.owns)) ∧ pre(gs, ls, D2, a) ∧ pcstepf(gs, ls, D2, a) = D3 → (∃ rl. gs.orl + gs.qrl + rl = gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl) ∧ (∃ rl0. gs.orl + rl0 = gstepf(gs, ls, D2, a).orl) ∧ gstepf(gs, ls, D2, a).vtail ∈ gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl from gs.vtail ∧ gstepf(gs, ls, D2, a).phead ∈ gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl from gs.phead ∧ gstepf(gs, ls, D2, a).vhead ∈ gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl from gs.vhead ∧ equalrl(gs.orl + gs.qrl, gs.vs, gs.ps, gstepf(gs, ls, D2, a).vs, gstepf(gs, ls, D2, a).ps) ∧ equalowns(gs.vs, gs.ps, gs.owns, tid0, gstepf(gs, ls, D2, a).vs, gstepf(gs, ls, D2, a).ps, gstepf(gs, ls, D2, a).owns) ∧ gs.vretRefs(tid0) = gstepf(gs, ls, D2, a).vretRefs(tid0) ∧ (gs.pretRefs(tid0) = gstepf(gs, ls, D2, a).pretRefs(tid0) ∨ gs.vretRefs(tid0) = gstepf(gs, ls, D2, a).pretRefs(tid0)) ∧ rvownsrely(tid0, gs.rvowns, gstepf(gs, ls, D2, a).rvowns, gs.vrvHp) ∧ rvrely(tid0, gs.rvowns, gs.vretRefs, gs.vrvHp, gstepf(gs, ls, D2, a).vretRefs, gstepf(gs, ls, D2, a).vrvHp) ∧ dom gs.vrvHp ⊆ dom gstepf(gs, ls, D2, a).vrvHp ∧ prvHpRely(gs.vrvHp, gs.prvHp, gstepf(gs, ls, D2, a).prvHp) ∧ noidRely(tid0, gs.orl + gs.qrl, gs.vs, gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl, gstepf(gs, ls, D2, a).vs) ∧ nextnottidRely(tid0, gs.vhead, gs.vs, gstepf(gs, ls, D2, a).vhead, gstepf(gs, ls, D2, a).vs) ∧ ownsRely(tid0, gs.owns, gstepf(gs, ls, D2, a).owns)

System applied rule normalize

And obtained the following new goals

Rule argument

Additional information

Proof step 2 from proof rely-D2-D3 (previous step) (back to tree)

dom gs.vrvHp = dom gs.prvHp, pcstepf(gs, ls, D2, a) = D3, gs.orl ≠ [], ls.tid ≠ tid0, allocrv(gs.vretRefs, gs.vrvHp), allocrv(gs.pretRefs, gs.prvHp), listp(gs.orl + gs.qrl, gs.vs, gs.ps, gs.owns), markedp(gs.orl, gs.ps), markedp(gs.orl, gs.vs), nextnottid(ls.tid, gs.vhead, gs.vs), noid(ls.tid, gs.orl + gs.qrl, gs.vhead, gs.vs), nonullRetRefs(gs.orl + gs.qrl, gs.vs, gs.vretRefs, gs.pretRefs), noowns(ls.tid, gs.owns), pre(gs, ls, D2, a), rvownerok(gs.vretRefs, gs.vrvHp, gs.rvowns), rvownerok(gs.pretRefs, gs.prvHp, gs.rvowns), unmarkedp(gs.qrl, gs.ps), gs.phead ∈ gs.orl, ¬ dups(gs.orl + gs.qrl), gs.vs[gs.vhead].next ≠ null ∧ gs.vs[gs.vs[gs.vhead].next].deqID ≠ None → gs.rvowns(gs.vretRefs(gs.vs[gs.vs[gs.vhead].next].deqID.tid)) = Claimed(gs.vs[gs.vs[gs.vhead].next].deqID.tid, gs.vs[gs.vs[gs.vhead].next].val), gs.orl.last = gs.vhead ∧ (gs.qrl ≠ [] → unmarkedp(gs.qrl.tail, gs.vs)) ∨ gs.orl.butlast ≠ [] ∧ gs.orl.butlast.last = gs.vhead ∧ gs.phead ≠ gs.orl.last ∧ unmarkedp(gs.qrl, gs.vs), gs.vtail = (gs.orl + gs.qrl).last ∨ gs.qrl ≠ [] ∧ gs.vtail = (gs.orl + gs.qrl.butlast).last ∧ gs.vs[gs.qrl.last].deqID = None ∧ gs.vs[(gs.orl + gs.qrl).last].next = null ⊦ (∃ rl. gs.orl + gs.qrl + rl = gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl) ∧ (∃ rl0. gs.orl + rl0 = gstepf(gs, ls, D2, a).orl) ∧ gstepf(gs, ls, D2, a).vtail ∈ gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl from gs.vtail ∧ gstepf(gs, ls, D2, a).phead ∈ gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl from gs.phead ∧ gstepf(gs, ls, D2, a).vhead ∈ gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl from gs.vhead ∧ equalrl(gs.orl + gs.qrl, gs.vs, gs.ps, gstepf(gs, ls, D2, a).vs, gstepf(gs, ls, D2, a).ps) ∧ equalowns(gs.vs, gs.ps, gs.owns, tid0, gstepf(gs, ls, D2, a).vs, gstepf(gs, ls, D2, a).ps, gstepf(gs, ls, D2, a).owns) ∧ gs.vretRefs(tid0) = gstepf(gs, ls, D2, a).vretRefs(tid0) ∧ (gs.pretRefs(tid0) = gstepf(gs, ls, D2, a).pretRefs(tid0) ∨ gs.vretRefs(tid0) = gstepf(gs, ls, D2, a).pretRefs(tid0)) ∧ rvownsrely(tid0, gs.rvowns, gstepf(gs, ls, D2, a).rvowns, gs.vrvHp) ∧ rvrely(tid0, gs.rvowns, gs.vretRefs, gs.vrvHp, gstepf(gs, ls, D2, a).vretRefs, gstepf(gs, ls, D2, a).vrvHp) ∧ dom gs.vrvHp ⊆ dom gstepf(gs, ls, D2, a).vrvHp ∧ prvHpRely(gs.vrvHp, gs.prvHp, gstepf(gs, ls, D2, a).prvHp) ∧ noidRely(tid0, gs.orl + gs.qrl, gs.vs, gstepf(gs, ls, D2, a).orl + gstepf(gs, ls, D2, a).qrl, gstepf(gs, ls, D2, a).vs) ∧ nextnottidRely(tid0, gs.vhead, gs.vs, gstepf(gs, ls, D2, a).vhead, gstepf(gs, ls, D2, a).vs) ∧ ownsRely(tid0, gs.owns, gstepf(gs, ls, D2, a).owns)

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: ⊦ s ⊆ s ++ a; (s) ⊦ ownsRely(tid, owns, owns); (s) ⊦ (gs.vrvHp:= vrvHp).owns = gs.owns; (s) ⊦ (gs.prvHp:= vrvHp0).owns = gs.owns; (s) ⊦ (gs.rvowns:= rvowns).owns = gs.owns; (s) ⊦ (ls.rvref:= rvref).tid = ls.tid; (s) ⊦ (ls.optval0:= optval0).tid = ls.tid; (s) ⊦ (ls.rvref:= rvref).rvref = rvref; (s) ⊦ (ls.optval0:= optval0).rvref = ls.rvref; (s) ⊦ (ls.optval0:= optval0).optval0 = optval0; (s) ⊦ gstepf(gs, ls, D2, chooseD2(rvref0, optval00)) = ( ( gs .vrvHp:= gs.vrvHp [((ls.rvref:= rvref0).optval0:= optval00 ).rvref, ((ls.rvref:= rvref0).optval0:= optval00 ).optval0 ]) .prvHp:= gs.prvHp [((ls.rvref:= rvref0).optval0:= optval00).rvref, ((ls.rvref:= rvref0).optval0:= optval00).optval0 ]) .rvowns:= gs.rvowns (((ls.rvref:= rvref0).optval0:= optval00).rvref ; Pending(((ls.rvref:= rvref0).optval0:= optval00).tid) ); (s) ⊦ nextnottidRely(tid, vhead, vs, vhead, vs); (s) ⊦ (gs.vrvHp:= vrvHp).vs = gs.vs; (s) ⊦ (gs.prvHp:= vrvHp0).vs = gs.vs; (s) ⊦ (gs.rvowns:= rvowns).vs = gs.vs; (s) ⊦ (gs.vrvHp:= vrvHp).vhead = gs.vhead; (s) ⊦ (gs.prvHp:= vrvHp0).vhead = gs.vhead; (s) ⊦ (gs.rvowns:= rvowns).vhead = gs.vhead; (s) ⊦ noidRely(tid, rl, vs, rl, vs); (s) ⊦ (gs.vrvHp:= vrvHp).qrl = gs.qrl; (s) ⊦ (gs.prvHp:= vrvHp0).qrl = gs.qrl; (s) ⊦ (gs.rvowns:= rvowns).qrl = gs.qrl; (s) ⊦ (gs.vrvHp:= vrvHp).orl = gs.orl; (s) ⊦ (gs.prvHp:= vrvHp0).orl = gs.orl; (s) ⊦ (gs.rvowns:= rvowns).orl = gs.orl; (s) ⊦ (gs.prvHp:= vrvHp0).prvHp = vrvHp0; (s) ⊦ (gs.rvowns:= rvowns).prvHp = gs.prvHp; (s) ⊦ dom st[a, d] = dom st ++ a; (s) ⊦ (gs.vrvHp:= vrvHp).vrvHp = vrvHp; (s) ⊦ (gs.prvHp:= vrvHp0).vrvHp = gs.vrvHp; (s) ⊦ (gs.rvowns:= rvowns).vrvHp = gs.vrvHp; (s) ⊦ (gs.vrvHp:= vrvHp).vretRefs = gs.vretRefs; (s) ⊦ (gs.prvHp:= vrvHp0).vretRefs = gs.vretRefs; (s) ⊦ (gs.rvowns:= rvowns).vretRefs = gs.vretRefs; (s) ⊦ (gs.rvowns:= rvowns).rvowns = rvowns; (s) ⊦ (gs.vrvHp:= vrvHp).pretRefs = gs.pretRefs; (s) ⊦ (gs.prvHp:= vrvHp0).pretRefs = gs.pretRefs; (s) ⊦ (gs.rvowns:= rvowns).pretRefs = gs.pretRefs; (s) ⊦ equalowns(vs, ps, owns, tid, vs, ps, owns); (s) ⊦ (gs.vrvHp:= vrvHp).ps = gs.ps; (s) ⊦ (gs.prvHp:= vrvHp0).ps = gs.ps; (s) ⊦ (gs.rvowns:= rvowns).ps = gs.ps; (s) ⊦ equalrl(rl, vs, ps, vs, ps); (s) a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ r ∈ rl → r ∈ rl from r; (ss) ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s) ⊦ r ∈ orl → orl + qrl from r = (orl from r) + qrl; (s) ⊦ (gs.vrvHp:= vrvHp).phead = gs.phead; (s) ⊦ (gs.prvHp:= vrvHp0).phead = gs.phead; (s) ⊦ (gs.rvowns:= rvowns).phead = gs.phead; (s) ⊦ (gs.vrvHp:= vrvHp).vtail = gs.vtail; (s) ⊦ (gs.prvHp:= vrvHp0).vtail = gs.vtail; (s) ⊦ (gs.rvowns:= rvowns).vtail = gs.vtail; (s) ⊦ x = x + y ↔ y = []; (s) ⊦ # x = # z → (x + y = z + y0 ↔ x = z ∧ y = y0); (ws) ⊦ chooseD2(rvref, optval) = chooseD2(rvref0, optval0) ↔ rvref = rvref0 ∧ optval = optval0; (ws) ⊦ chooseD2(rvref, optval) ≠ τ; (s) a: ⊦ ¬ dups(x + y) → ¬ dups(x) ∧ ¬ dups(y); (forward) from specification list-dup[reflist] list-elem-in-app11: ⊦ listp(orl + qrl, vs, ps, owns) ∧ node ∈ orl → node ∈ vs; (forward) from specification DMSQ list-elem-in-app12: ⊦ listp(orl + qrl, vs, ps, owns) ∧ node ∈ orl → node ∈ ps; (forward) from specification DMSQ ⊦ y ≠ [] → (x + y).last = y.last; (s) ⊦ pre(gs, ls, D2, a) ↔ (∃ rvref0, optval00. (¬ rvref0 ∈ gs.vrvHp ∧ ¬ rvref0 ∈ gs.prvHp ∧ rvref0 ≠ null) ∧ a = chooseD2(rvref0, optval00)) ∨ a = τ ∧ ¬ (∃ rvref0, optval00. ¬ rvref0 ∈ gs.vrvHp ∧ ¬ rvref0 ∈ gs.prvHp ∧ rvref0 ≠ null); (s) ⊦ D2 ≠ D3; (s) ⊦ pcstepf(gs, ls, D2, a) = (∃ rvref0, optval00. a = chooseD2(rvref0, optval00) ⊃ D3;D2); (s)

Proof step 3 from proof rely-D2-D3 (previous step) (back to tree)

dom gs.vrvHp = dom gs.prvHp, gs.orl ≠ [], ls.tid ≠ tid0, rvref ≠ null, allocrv(gs.pretRefs, gs.prvHp), allocrv(gs.vretRefs, gs.vrvHp), listp(gs.orl + gs.qrl, gs.vs, gs.ps, gs.owns), markedp(gs.orl, gs.vs), markedp(gs.orl, gs.ps), nextnottid(ls.tid, gs.vhead, gs.vs), noid(ls.tid, gs.orl + gs.qrl, gs.vhead, gs.vs), nonullRetRefs(gs.orl + gs.qrl, gs.vs, gs.vretRefs, gs.pretRefs), noowns(ls.tid, gs.owns), rvownerok(gs.pretRefs, gs.prvHp, gs.rvowns), rvownerok(gs.vretRefs, gs.vrvHp, gs.rvowns), unmarkedp(gs.qrl, gs.ps), gs.phead ∈ gs.orl, gs.phead ∈ gs.vs, gs.phead ∈ gs.ps, ¬ dups(gs.orl + gs.qrl), ¬ dups(gs.orl), ¬ dups(gs.qrl), ¬ rvref ∈ gs.vrvHp, ¬ rvref ∈ gs.prvHp, gs.vs[gs.vhead].next ≠ null ∧ gs.vs[gs.vs[gs.vhead].next].deqID ≠ None → gs.rvowns(gs.vretRefs(gs.vs[gs.vs[gs.vhead].next].deqID.tid)) = Claimed(gs.vs[gs.vs[gs.vhead].next].deqID.tid, gs.vs[gs.vs[gs.vhead].next].val), gs.vtail = (gs.orl + gs.qrl).last ∨ gs.qrl ≠ [] ∧ gs.vtail = (gs.orl + gs.qrl.butlast).last ∧ gs.vs[gs.qrl.last].deqID = None ∧ gs.vs[gs.qrl.last].next = null, gs.orl.last = gs.vhead ∧ (gs.qrl ≠ [] → unmarkedp(gs.qrl.tail, gs.vs)) ∨ gs.orl.butlast ≠ [] ∧ gs.orl.butlast.last = gs.vhead ∧ gs.phead ≠ gs.orl.last ∧ unmarkedp(gs.qrl, gs.vs) ⊦ gs.vtail ∈ gs.orl + gs.qrl from gs.vtail ∧ gs.vhead ∈ gs.orl + gs.qrl from gs.vhead ∧ rvownsrely(tid0, gs.rvowns, gs.rvowns(rvref; Pending(ls.tid)), gs.vrvHp) ∧ rvrely(tid0, gs.rvowns, gs.vretRefs, gs.vrvHp, gs.vretRefs, gs.vrvHp[rvref, optval00]) ∧ prvHpRely(gs.vrvHp, gs.prvHp, gs.prvHp[rvref, optval00])

Heuristic elimination applied rule apply elim lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:elim-mkgs, seq:⊦ ( ( ( ( ( ( ( ( ( ( (ps = gs.ps ∧ vs = gs.vs) ∧ phead = gs.phead) ∧ vhead = gs.vhead) ∧ vtail = gs.vtail) ∧ vretRefs = gs.vretRefs) ∧ pretRefs = gs.pretRefs) ∧ vrvHp = gs.vrvHp) ∧ prvHp = gs.prvHp) ∧ qrl = gs.qrl) ∧ orl = gs.orl) ∧ owns = gs.owns) ∧ rvowns = gs.rvowns ↔ gs = mkgs (ps, vs, phead, vhead, vtail, vretRefs, pretRefs, vrvHp, prvHp, qrl, orl, owns, rvowns), subst:[qrl, ps, phead, vrvHp, vretRefs, vhead, vtail, gs, owns, rvowns, orl, prvHp, pretRefs, vs] → [qrl, ps, phead, vrvHp, vretRefs, vhead, vtail, gs, owns, rvowns, orl, prvHp, pretRefs, vs]

Additional information

Used simplifier rules: ⊦ ( ( ( ( ( ( ( ( ( ((ps = gs.ps ∧ vs = gs.vs) ∧ phead = gs.phead) ∧ vhead = gs.vhead) ∧ vtail = gs.vtail) ∧ vretRefs = gs.vretRefs) ∧ pretRefs = gs.pretRefs) ∧ vrvHp = gs.vrvHp) ∧ prvHp = gs.prvHp) ∧ qrl = gs.qrl) ∧ orl = gs.orl) ∧ owns = gs.owns) ∧ rvowns = gs.rvowns ↔ gs = mkgs (ps, vs, phead, vhead, vtail, vretRefs, pretRefs, vrvHp, prvHp, qrl, orl, owns, rvowns); (elim) Used terms: qrl, ps, phead, vrvHp, vretRefs, vhead, vtail, gs, owns, rvowns, orl, prvHp, pretRefs, vs

Proof step 4 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, gs = mkgs(ps, vs, phead, vhead, vtail, vretRefs, pretRefs, vrvHp, prvHp, qrl, orl, owns, rvowns), ls.tid ≠ tid0, orl ≠ [], rvref ≠ null, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(ls.tid, vhead, vs), noid(ls.tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(ls.tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦ vtail ∈ orl + qrl from vtail ∧ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(ls.tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] a: ⊦ ¬ dups(x + y) → ¬ dups(x) ∧ ¬ dups(y); (forward) from specification list-dup[reflist] list-elem-in-app11: ⊦ listp(orl + qrl, vs, ps, owns) ∧ node ∈ orl → node ∈ vs; (forward) from specification DMSQ list-elem-in-app12: ⊦ listp(orl + qrl, vs, ps, owns) ∧ node ∈ orl → node ∈ ps; (forward) from specification DMSQ

Proof step 5 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, ls.tid ≠ tid0, orl ≠ [], rvref ≠ null, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(ls.tid, vhead, vs), noid(ls.tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(ls.tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦ vtail ∈ orl + qrl from vtail ∧ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(ls.tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

Heuristic elimination applied rule apply elim lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:elim-mkls, seq:⊦ ( ( ( ( ( ( ( ( ( ( (tid = ls.tid ∧ value = ls.value) ∧ optval = ls.optval) ∧ node = ls.node) ∧ nd = ls.nd) ∧ (success ↔ ls.success)) ∧ lst = ls.lst) ∧ nxt = ls.nxt) ∧ rvref = ls.rvref) ∧ optval0 = ls.optval0) ∧ first = ls.first) ∧ othertid = ls.othertid) ∧ addr = ls.addr ↔ ls = mkls (tid, value, optval, node, nd, success, lst, nxt, rvref, optval0, first, othertid, addr), subst:[value, nxt, optval0, node, nd, ls, addr, tid, lst, first, rvref, othertid, success, optval] → [value, nxt, optval0, node, nd, ls, addr, tid, lst, first, rvref0, othertid, success, optval]

Additional information

Used simplifier rules: ⊦ ( ( ( ( ( ( ( ( ( ( (tid = ls.tid ∧ value = ls.value) ∧ optval = ls.optval) ∧ node = ls.node) ∧ nd = ls.nd) ∧ (success ↔ ls.success)) ∧ lst = ls.lst) ∧ nxt = ls.nxt) ∧ rvref = ls.rvref) ∧ optval0 = ls.optval0) ∧ first = ls.first) ∧ othertid = ls.othertid) ∧ addr = ls.addr ↔ ls = mkls (tid, value, optval, node, nd, success, lst, nxt, rvref, optval0, first, othertid, addr); (elim) Used terms: value, nxt, optval, node, nd, ls, addr, tid, lst, first, rvref0, othertid, success, optval

Proof step 6 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, ls = mkls(tid, value, optval, node, nd, success, lst, nxt, rvref0, optval0, first, othertid, addr), orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦ vtail ∈ orl + qrl from vtail ∧ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Proof step 7 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦ vtail ∈ orl + qrl from vtail ∧ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:from-in2, seq: ⊦ r ∈ rl → r ∈ rl from r, subst:[r, rl] → [vtail, orl + qrl], <all paths>

Additional information

Used terms: vtail, orl + qrl

Proof step 8 from proof rely-D2-D3 (previous step) (back to tree)

⊦ r ∈ rl → r ∈ rl from r

A spec-lemma

Proof step 9 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ vtail ∈ orl + qrl, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦ vtail ∈ orl + qrl from vtail ∧ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: ⊦ ¬ a ∈ []; (s) a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ ¬ r ∈ rl → rl from r = []; (s) ⊦ ¬ r ∈ orl → orl + qrl from r = qrl from r; (s) ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s)

Proof step 10 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ vtail ∈ orl, ¬ vtail ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦

User applied rule case distinction

And obtained the following new goals

Rule argument

pos: left 28

Additional information

Proof step 11 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, vtail = (orl + qrl).last, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ vtail ∈ orl, ¬ vtail ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Proof step 12 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (orl + qrl).last ∈ orl, ¬ (orl + qrl).last ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: list-last[reflist], name:lapp, seq: ⊦ y ≠ [] → (x + y).last = y.last, subst:[x, y] → [orl, qrl], <all paths>

Additional information

Used simplifier rules: ⊦ x.last ∈ x ↔ x ≠ []; (s) ⊦ unmarkedp([], vs); (s) ⊦ ¬ dups([]); (s) ⊦ x + [] = x; (s) a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] Used terms: orl, qrl

Proof step 13 from proof rely-D2-D3 (previous step) (back to tree)

⊦ y ≠ [] → (x + y).last = y.last

A spec-lemma

Proof step 14 from proof rely-D2-D3 (previous step) (back to tree)

vs[qrl.last].deqID = None, vs[qrl.last].next = null, dom vrvHp = dom prvHp, vtail = (orl + qrl.butlast).last, orl ≠ [], qrl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ vtail ∈ orl, ¬ vtail ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Proof step 15 from proof rely-D2-D3 (previous step) (back to tree)

vs[qrl.last].deqID = None, vs[qrl.last].next = null, dom vrvHp = dom prvHp, orl ≠ [], qrl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (orl + qrl.butlast).last ∈ orl, ¬ (orl + qrl.butlast).last ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ unmarkedp(qrl.tail, vs) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦

Heuristic elimination applied rule apply elim lemma

And obtained the following new goals

Rule argument

spec: list-data[reflist], name:elim-add, seq: ⊦ x ≠ [] → (head = x.head ∧ tail = x.tail ↔ x = head + tail), subst:[x, head, tail] → [qrl, phead0, orl0]

Additional information

Used simplifier rules: ⊦ x ≠ [] → (head = x.head ∧ tail = x.tail ↔ x = head + tail); (elim) Used terms: qrl, head, tail

Proof step 16 from proof rely-D2-D3 (previous step) (back to tree)

vs[qrl.last].deqID = None, vs[qrl.last].next = null, dom vrvHp = dom prvHp, qrl = phead0 + orl0, orl ≠ [], qrl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (orl + qrl.butlast).last ∈ orl, ¬ (orl + qrl.butlast).last ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ unmarkedp(orl0, vs) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs) ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] a: ⊦ ¬ dups(x + y) → ¬ dups(x) ∧ ¬ dups(y); (forward) from specification list-dup[reflist] ⊦ unmarkedp(r ', vs) ↔ vs[r].deqID = None; (s) ⊦ unmarkedp(rl + rl0, vs) ↔ unmarkedp(rl, vs) ∧ unmarkedp(rl0, vs); (s) ⊦ a + x = a ' + x; (s) ⊦ a ∈ b ' + x ↔ ¬(a ≠ b ∧ ¬ a ∈ x); (ss) ⊦ dups(a ' + x) ↔ a ∈ x ∨ dups(x); (s) ⊦ dups(x + a ' + y) ↔ a ∈ x ∨ a ∈ y ∨ dups(x + y); (ws) ⊦ a ' + x ≠ []; (s)

Proof step 17 from proof rely-D2-D3 (previous step) (back to tree)

vs[(phead0 ' + orl0).last].deqID = None, ps[phead0].deqID = None, vs[(phead0 ' + orl0).last].next = null, dom vrvHp = dom prvHp, (orl + (phead0 ' + orl0).butlast).last ≠ phead0, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + phead0 ' + orl0, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + phead0 ' + orl0, vhead, vs), nonullRetRefs(orl + phead0 ' + orl0, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(orl0, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + orl0), ¬ dups(orl), ¬ dups(orl0), ¬ phead0 ∈ orl, ¬ phead0 ∈ orl0, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (orl + (phead0 ' + orl0).butlast).last ∈ orl, ¬ (orl + (phead0 ' + orl0).butlast).last ∈ orl0, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ unmarkedp(orl0, vs) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ vs[phead0].deqID = None ∧ unmarkedp(orl0, vs) ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: list-last[reflist], name:bapp, seq: ⊦ y ≠ [] → (x + y).butlast = x + y.butlast, subst:[x, y] → [phead0 ', orl0], <all paths>

Additional information

Used simplifier rules: ⊦ a '.last = a; (s) ⊦ unmarkedp([], vs); (s) ⊦ ¬ dups([]); (s) ⊦ ¬ a ∈ []; (s) ⊦ a '.butlast = []; (s) ⊦ x + [] = x; (s) a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ x.last ∈ x ↔ x ≠ []; (s) Used terms: phead0 ', orl0

Proof step 18 from proof rely-D2-D3 (previous step) (back to tree)

⊦ y ≠ [] → (x + y).butlast = x + y.butlast

A spec-lemma

Proof step 19 from proof rely-D2-D3 (previous step) (back to tree)

ps[phead0].deqID = None, vs[(phead0 ' + orl0).last].deqID = None, vs[(phead0 ' + orl0).last].next = null, dom vrvHp = dom prvHp, (orl + phead0 ' + orl0.butlast).last ≠ phead0, orl ≠ [], orl0 ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + phead0 ' + orl0, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + phead0 ' + orl0, vhead, vs), nonullRetRefs(orl + phead0 ' + orl0, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(orl0, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + orl0), ¬ dups(orl), ¬ dups(orl0), ¬ phead0 ∈ orl, ¬ phead0 ∈ orl0, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (orl + phead0 ' + orl0.butlast).last ∈ orl, ¬ (orl + phead0 ' + orl0.butlast).last ∈ orl0, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ unmarkedp(orl0, vs) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ vs[phead0].deqID = None ∧ unmarkedp(orl0, vs) ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ (x + a ' + y).last = (a ' + y).last; (s) ⊦ x ≠ [] → (a ' + x).last = x.last; (s)

Proof step 20 from proof rely-D2-D3 (previous step) (back to tree)

ps[phead0].deqID = None, vs[orl0.last].deqID = None, vs[orl0.last].next = null, dom vrvHp = dom prvHp, (phead0 ' + orl0.butlast).last ≠ phead0, orl ≠ [], orl0 ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + phead0 ' + orl0, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + phead0 ' + orl0, vhead, vs), nonullRetRefs(orl + phead0 ' + orl0, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(orl0, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + orl0), ¬ dups(orl), ¬ dups(orl0), ¬ phead0 ∈ orl, ¬ phead0 ∈ orl0, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (phead0 ' + orl0.butlast).last ∈ orl, ¬ (phead0 ' + orl0.butlast).last ∈ orl0, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ unmarkedp(orl0, vs) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ vs[phead0].deqID = None ∧ unmarkedp(orl0, vs) ⊦

User applied rule apply elim lemma

And obtained the following new goals

Rule argument

spec: list-last[reflist], name:last-elim, seq:x ≠ [] ⊦ a = x.last ∧ y = x.butlast ↔ x = y + a ', subst:[x, a, y] → [orl0, vhead0, orl1]

Additional information

Used simplifier rules: x ≠ [] ⊦ a = x.last ∧ y = x.butlast ↔ x = y + a '; (elim) Used terms: orl0, vhead0, y

Proof step 21 from proof rely-D2-D3 (previous step) (back to tree)

vs[vhead0].deqID = None, ps[phead0].deqID = None, vs[vhead0].next = null, dom vrvHp = dom prvHp, orl0 = orl1 + vhead0 ', (phead0 ' + orl1).last ≠ phead0, orl ≠ [], orl0 ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + phead0 ' + orl0, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + phead0 ' + orl0, vhead, vs), nonullRetRefs(orl + phead0 ' + orl0, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(orl0, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + orl0), ¬ dups(orl), ¬ dups(orl0), ¬ phead0 ∈ orl, ¬ phead0 ∈ orl0, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (phead0 ' + orl1).last ∈ orl, ¬ (phead0 ' + orl1).last ∈ orl0, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ unmarkedp(orl0, vs) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ vs[phead0].deqID = None ∧ unmarkedp(orl0, vs) ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] a: ⊦ ¬ dups(x + y) → ¬ dups(x) ∧ ¬ dups(y); (forward) from specification list-dup[reflist] ⊦ unmarkedp(r ', vs) ↔ vs[r].deqID = None; (s) ⊦ unmarkedp(rl + rl0, vs) ↔ unmarkedp(rl, vs) ∧ unmarkedp(rl0, vs); (s) ⊦ a ∈ b ' ↔ a = b; (s) ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s) ⊦ ¬ b ∈ x → (dups(x + b ') ↔ dups(x)); (s) ⊦ dups(x + a ') ↔ a ∈ x ∨ dups(x); (ws) ⊦ x + a ' ≠ []; (s)

Proof step 22 from proof rely-D2-D3 (previous step) (back to tree)

ps[phead0].deqID = None, vs[vhead0].deqID = None, ps[vhead0].deqID = None, vs[vhead0].next = null, dom vrvHp = dom prvHp, (phead0 ' + orl1).last ≠ phead0, (phead0 ' + orl1).last ≠ vhead0, orl ≠ [], phead0 ≠ vhead0, rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + phead0 ' + orl1 + vhead0 ', vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + phead0 ' + orl1 + vhead0 ', vhead, vs), nonullRetRefs(orl + phead0 ' + orl1 + vhead0 ', vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(orl1, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + orl1), ¬ dups(orl), ¬ dups(orl1), ¬ vhead0 ∈ orl, ¬ vhead0 ∈ orl1, ¬ phead0 ∈ orl, ¬ phead0 ∈ orl1, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (phead0 ' + orl1).last ∈ orl, ¬ (phead0 ' + orl1).last ∈ orl1, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ unmarkedp(orl1, vs) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ vs[phead0].deqID = None ∧ unmarkedp(orl1, vs) ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: list-last[reflist], name:lapp, seq: ⊦ y ≠ [] → (x + y).last = y.last, subst:[x, y] → [phead0 ', orl1], <all paths>

Additional information

Used simplifier rules: ⊦ x.last ∈ x ↔ x ≠ []; (s) ⊦ x + [] = x; (s) a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ a '.last = a; (s) Used terms: phead0 ', orl1

Proof step 23 from proof rely-D2-D3 (previous step) (back to tree)

⊦ y ≠ [] → (x + y).last = y.last

A spec-lemma

Proof step 24 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, vtail ∈ orl + qrl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦ true ∧ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s)

Proof step 25 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail ∈ orl ∨ vtail ∈ qrl ⊦ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:from-in2, seq: ⊦ r ∈ rl → r ∈ rl from r, subst:[r, rl] → [vhead, orl + qrl], <all paths>

Additional information

Used terms: vhead, orl + qrl

Proof step 26 from proof rely-D2-D3 (previous step) (back to tree)

⊦ r ∈ rl → r ∈ rl from r

A spec-lemma

Proof step 27 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ vhead ∈ orl + qrl, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail ∈ orl ∨ vtail ∈ qrl, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦ vhead ∈ orl + qrl from vhead ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: ⊦ ¬ a ∈ []; (s) a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ ¬ r ∈ rl → rl from r = []; (s) ⊦ ¬ r ∈ orl → orl + qrl from r = qrl from r; (s) ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s)

Proof step 28 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ vhead ∈ orl, ¬ vhead ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail ∈ orl ∨ vtail ∈ qrl ⊦

User applied rule case distinction

And obtained the following new goals

Rule argument

pos: left 29

Additional information

Used simplifier rules: ⊦ x.last ∈ x ↔ x ≠ []; (s)

Proof step 29 from proof rely-D2-D3 (previous step) (back to tree)

orl.butlast.last = vhead, dom vrvHp = dom prvHp, orl.butlast ≠ [], orl ≠ [], phead ≠ orl.last, rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), unmarkedp(qrl, vs), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ vhead ∈ orl, ¬ vhead ∈ qrl, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail ∈ orl ∨ vtail ∈ qrl, vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: ⊦ x ≠ [] → (x.butlast = [] ↔ x = x.head '); (ss)

Proof step 30 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], orl ≠ orl.head ', phead ≠ orl.last, rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, orl.butlast.last, vs), noid(tid, orl + qrl, orl.butlast.last, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, vs), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ orl.butlast.last ∈ orl, ¬ orl.butlast.last ∈ qrl, vs[orl.butlast.last].next ≠ null ∧ vs[vs[orl.butlast.last].next].deqID ≠ None → rvowns(vretRefs(vs[vs[orl.butlast.last].next].deqID.tid)) = Claimed(vs[vs[orl.butlast.last].next].deqID.tid, vs[vs[orl.butlast.last].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, vtail ∈ orl ∨ vtail ∈ qrl ⊦

Heuristic elimination applied rule apply elim lemma

And obtained the following new goals

Rule argument

spec: list-data[reflist], name:elim-add, seq: ⊦ x ≠ [] → (head = x.head ∧ tail = x.tail ↔ x = head + tail), subst:[x, head, tail] → [orl, vtail0, qrl0]

Additional information

Used simplifier rules: ⊦ x ≠ [] → (head = x.head ∧ tail = x.tail ↔ x = head + tail); (elim) Used terms: orl, head, tail

Proof step 31 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl = vtail0 + qrl0, orl ≠ [], orl ≠ vtail0 ', phead ≠ orl.last, rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, orl.butlast.last, vs), noid(tid, orl + qrl, orl.butlast.last, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), unmarkedp(qrl, vs), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ orl.butlast.last ∈ orl, ¬ orl.butlast.last ∈ qrl, vs[orl.butlast.last].next ≠ null ∧ vs[vs[orl.butlast.last].next].deqID ≠ None → rvowns(vretRefs(vs[vs[orl.butlast.last].next].deqID.tid)) = Claimed(vs[vs[orl.butlast.last].next].deqID.tid, vs[vs[orl.butlast.last].next].val), vtail ∈ orl ∨ vtail ∈ qrl, vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] a: ⊦ ¬ dups(x + y) → ¬ dups(x) ∧ ¬ dups(y); (forward) from specification list-dup[reflist] ⊦ x ≠ [] → (a ' + x).butlast = a ' + x.butlast; (s) ⊦ a + x = a ' + x; (s) ⊦ a ∈ b ' + x ↔ a = b ∨ a ∈ x; (s) ⊦ y ≠ [] → (x + y + z).last = (y + z).last; (s) ⊦ a ∈ b ' + x ↔ ¬(a ≠ b ∧ ¬ a ∈ x); (ss) ⊦ dups(a ' + x) ↔ a ∈ x ∨ dups(x); (s) ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s) ⊦ markedp(r ', vs) ↔ some?(vs[r].deqID); (s) ⊦ markedp(rl + rl0, vs) ↔ markedp(rl, vs) ∧ markedp(rl0, vs); (s) ⊦ x ≠ [] → (a ' + x).last = x.last; (s) ⊦ a ' = b ' + x ↔ a = b ∧ x = []; (ws) ⊦ a ' + x ≠ []; (s)

Proof step 32 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, (vtail0 ' + qrl0.butlast).last ≠ vtail0, phead ≠ qrl0.last, qrl0 ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp((vtail0 ' + qrl0) + qrl, vs, ps, owns), markedp(qrl0, vs), markedp(qrl0, ps), nextnottid(tid, (vtail0 ' + qrl0.butlast).last, vs), noid(tid, (vtail0 ' + qrl0) + qrl, (vtail0 ' + qrl0.butlast).last, vs), nonullRetRefs((vtail0 ' + qrl0) + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), some?(vs[vtail0].deqID), some?(ps[vtail0].deqID), unmarkedp(qrl, vs), unmarkedp(qrl, ps), phead ∈ vs, phead ∈ ps, ¬ dups(qrl0 + qrl), ¬ dups(qrl0), ¬ dups(qrl), ¬ vtail0 ∈ qrl0, ¬ vtail0 ∈ qrl, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (vtail0 ' + qrl0.butlast).last ∈ qrl0, ¬ (vtail0 ' + qrl0.butlast).last ∈ qrl, vs[(vtail0 ' + qrl0.butlast).last].next ≠ null ∧ vs[vs[(vtail0 ' + qrl0.butlast).last].next].deqID ≠ None → rvowns(vretRefs(vs[vs[(vtail0 ' + qrl0.butlast).last].next].deqID.tid)) = Claimed(vs[vs[(vtail0 ' + qrl0.butlast).last].next].deqID.tid, vs[vs[(vtail0 ' + qrl0.butlast).last].next].val), phead = vtail0 ∨ phead ∈ qrl0, vtail = (qrl0 + qrl).last ∨ qrl ≠ [] ∧ vtail = (qrl0 + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, vtail = vtail0 ∨ vtail ∈ qrl0 ∨ vtail ∈ qrl ⊦

User applied rule apply elim lemma

And obtained the following new goals

Rule argument

spec: list-last[reflist], name:last-elim, seq:x ≠ [] ⊦ a = x.last ∧ y = x.butlast ↔ x = y + a ', subst:[x, a, y] → [qrl0, phead0, qrl1]

Additional information

Used simplifier rules: x ≠ [] ⊦ a = x.last ∧ y = x.butlast ↔ x = y + a '; (elim) Used terms: qrl0, phead0, y

Proof step 33 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, qrl0 = qrl1 + phead0 ', (vtail0 ' + qrl1).last ≠ vtail0, phead ≠ phead0, qrl0 ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp((vtail0 ' + qrl0) + qrl, vs, ps, owns), markedp(qrl0, ps), markedp(qrl0, vs), nextnottid(tid, (vtail0 ' + qrl1).last, vs), noid(tid, (vtail0 ' + qrl0) + qrl, (vtail0 ' + qrl1).last, vs), nonullRetRefs((vtail0 ' + qrl0) + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), some?(ps[vtail0].deqID), some?(vs[vtail0].deqID), unmarkedp(qrl, ps), unmarkedp(qrl, vs), phead ∈ ps, phead ∈ vs, ¬ dups(qrl0 + qrl), ¬ dups(qrl0), ¬ dups(qrl), ¬ vtail0 ∈ qrl0, ¬ vtail0 ∈ qrl, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (vtail0 ' + qrl1).last ∈ qrl0, ¬ (vtail0 ' + qrl1).last ∈ qrl, vs[(vtail0 ' + qrl1).last].next ≠ null ∧ vs[vs[(vtail0 ' + qrl1).last].next].deqID ≠ None → rvowns(vretRefs(vs[vs[(vtail0 ' + qrl1).last].next].deqID.tid)) = Claimed(vs[vs[(vtail0 ' + qrl1).last].next].deqID.tid, vs[vs[(vtail0 ' + qrl1).last].next].val), vtail = vtail0 ∨ vtail ∈ qrl0 ∨ vtail ∈ qrl, vtail = (qrl0 + qrl).last ∨ qrl ≠ [] ∧ vtail = (qrl0 + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, phead = vtail0 ∨ phead ∈ qrl0 ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] a: ⊦ ¬ dups(x + y) → ¬ dups(x) ∧ ¬ dups(y); (forward) from specification list-dup[reflist] ⊦ a ∈ b ' ↔ a = b; (s) ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s) ⊦ (x + a ' + y).last = (a ' + y).last; (s) ⊦ ¬ b ∈ x → (dups(x + b ') ↔ dups(x)); (s) ⊦ dups(x + a ' + y) ↔ a ∈ x ∨ a ∈ y ∨ dups(x + y); (ws) ⊦ markedp(r ', vs) ↔ some?(vs[r].deqID); (s) ⊦ markedp(rl + rl0, vs) ↔ markedp(rl, vs) ∧ markedp(rl0, vs); (s) ⊦ x + a ' ≠ []; (s)

Proof step 34 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, (vtail0 ' + qrl1).last ≠ vtail0, (vtail0 ' + qrl1).last ≠ phead0, phead ≠ phead0, rvref ≠ null, tid ≠ tid0, vtail0 ≠ phead0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp((vtail0 ' + qrl1 + phead0 ') + qrl, vs, ps, owns), markedp(qrl1, vs), markedp(qrl1, ps), nextnottid(tid, (vtail0 ' + qrl1).last, vs), noid(tid, (vtail0 ' + qrl1 + phead0 ') + qrl, (vtail0 ' + qrl1).last, vs), nonullRetRefs((vtail0 ' + qrl1 + phead0 ') + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), some?(vs[vtail0].deqID), some?(ps[vtail0].deqID), some?(vs[phead0].deqID), some?(ps[phead0].deqID), unmarkedp(qrl, vs), unmarkedp(qrl, ps), phead ∈ vs, phead ∈ ps, ¬ dups(qrl1 + qrl), ¬ dups(qrl1), ¬ dups(qrl), ¬ (vtail0 ' + qrl1).last ∈ qrl, ¬ phead0 ∈ qrl1, ¬ phead0 ∈ qrl, ¬ vtail0 ∈ qrl1, ¬ vtail0 ∈ qrl, ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, ¬ (vtail0 ' + qrl1).last ∈ qrl1, vs[(vtail0 ' + qrl1).last].next ≠ null ∧ vs[vs[(vtail0 ' + qrl1).last].next].deqID ≠ None → rvowns(vretRefs(vs[vs[(vtail0 ' + qrl1).last].next].deqID.tid)) = Claimed(vs[vs[(vtail0 ' + qrl1).last].next].deqID.tid, vs[vs[(vtail0 ' + qrl1).last].next].val), phead = vtail0 ∨ phead ∈ qrl1, vtail = (phead0 ' + qrl).last ∨ qrl ≠ [] ∧ vtail = (phead0 ' + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, vtail = vtail0 ∨ vtail ∈ qrl1 ∨ vtail = phead0 ∨ vtail ∈ qrl ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: list-last[reflist], name:lapp, seq: ⊦ y ≠ [] → (x + y).last = y.last, subst:[x, y] → [vtail0 ', qrl1], <all paths>

Additional information

Used simplifier rules: ⊦ x.last ∈ x ↔ x ≠ []; (s) ⊦ x + [] = x; (s) a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ a '.last = a; (s) Used terms: vtail0 ', qrl1

Proof step 35 from proof rely-D2-D3 (previous step) (back to tree)

⊦ y ≠ [] → (x + y).last = y.last

A spec-lemma

Proof step 36 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, vhead ∈ orl + qrl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail ∈ orl ∨ vtail ∈ qrl, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦ true ∧ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: a: ⊦ (x + y) + z = x + y + z; from specification list[reflist] ⊦ a ∈ x + y ↔ a ∈ x ∨ a ∈ y; (s)

Proof step 37 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail ∈ orl ∨ vtail ∈ qrl, vhead ∈ orl ∨ vhead ∈ qrl ⊦ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp) ∧ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]) ∧ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00])

User applied rule case distinction

And obtained the following new goals

Rule argument

pos: right 1

Additional information

Proof step 38 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvownsrely(tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vhead ∈ orl ∨ vhead ∈ qrl, vtail ∈ orl ∨ vtail ∈ qrl, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:rvownsrely-def, seq:⊦ rvownsrely(tid, rvowns, rvowns0, vrvHp) ↔ (∀ rvref. rvref ∈ vrvHp → rvowns(rvref) = rvowns0(rvref) ∨ tid ≠ rvowns(rvref).tid ∧ rvowns(rvref).tid = rvowns0(rvref).tid ∧ rvowns(rvref).pending?), subst:[tid, rvowns, rvowns0, vrvHp] → [tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp], <all paths>

Additional information

Used terms: tid0, rvowns, rvowns(rvref; Pending(tid)), vrvHp

Proof step 39 from proof rely-D2-D3 (previous step) (back to tree)

⊦ rvownsrely(tid, rvowns, rvowns0, vrvHp) ↔ (∀ rvref. rvref ∈ vrvHp → rvowns(rvref) = rvowns0(rvref) ∨ tid ≠ rvowns(rvref).tid ∧ rvowns(rvref).tid = rvowns0(rvref).tid ∧ rvowns(rvref).pending?)

A spec-lemma

Proof step 40 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvrely(tid0, rvowns, vretRefs, vrvHp, vretRefs, vrvHp[rvref, optval00]), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vhead ∈ orl ∨ vhead ∈ qrl, vtail ∈ orl ∨ vtail ∈ qrl, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦

Heuristic use patterns applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:rvrely-def, seq:⊦ rvrely(tid, rvowns, vretRefs, vrvHp, vretRefs0, vrvHp0) ↔ vrvHp[vretRefs(tid)] = vrvHp0[vretRefs0(tid)] ∨ ¬ rvowns(vretRefs(tid)).pending? ∧ vrvHp0[vretRefs0(tid)] = Val(rvowns(vretRefs(tid)).val), subst:[vrvHp, vretRefs, tid, rvowns, vrvHp0, vretRefs0] → [vrvHp, vretRefs, tid0, rvowns, vrvHp[rvref, optval00], vretRefs], <all paths>

Additional information

Used terms: vrvHp, vretRefs, tid0, rvowns, vrvHp[rvref, optval00], vretRefs

Proof step 41 from proof rely-D2-D3 (previous step) (back to tree)

⊦ rvrely(tid, rvowns, vretRefs, vrvHp, vretRefs0, vrvHp0) ↔ vrvHp[vretRefs(tid)] = vrvHp0[vretRefs0(tid)] ∨ ¬ rvowns(vretRefs(tid)).pending? ∧ vrvHp0[vretRefs0(tid)] = Val(rvowns(vretRefs(tid)).val)

A spec-lemma

Proof step 42 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, vrvHp[vretRefs(tid0)] ≠ vrvHp[rvref, optval00][vretRefs(tid0)], orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail ∈ orl ∨ vtail ∈ qrl, vhead ∈ orl ∨ vhead ∈ qrl, rvowns(vretRefs(tid0)).pending? ∨ vrvHp[rvref, optval00][vretRefs(tid0)] ≠ Val(rvowns(vretRefs(tid0)).val) ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: store[retval-heap], name:At-other, seq: ⊦ a ≠ b → st[b, d][a] = st[a], subst:[st, b, d, a] → [vrvHp, rvref, optval00, vretRefs(tid0)], <all paths>

Additional information

Used terms: vrvHp, rvref, optval00, vretRefs(tid0)

Proof step 43 from proof rely-D2-D3 (previous step) (back to tree)

⊦ a ≠ b → st[b, d][a] = st[a]

A spec-lemma

Proof step 44 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, vretRefs(tid0) = rvref, vrvHp[vretRefs(tid0)] ≠ vrvHp[rvref, optval00][vretRefs(tid0)], orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), rvowns(vretRefs(tid0)).pending? ∨ vrvHp[rvref, optval00][vretRefs(tid0)] ≠ Val(rvowns(vretRefs(tid0)).val), vhead ∈ orl ∨ vhead ∈ qrl, vtail ∈ orl ∨ vtail ∈ qrl, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦

Heuristic simplifier applied rule simplifier

And obtained the following new goals

Rule argument

Additional information

Used simplifier rules: ⊦ a = b → st[b, d][a] = d; (s) ⊦ ¬ a ∈ st → st[a] = ∅[a]; (s)

Proof step 45 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, ∅[vretRefs(tid0)] ≠ optval00, orl ≠ [], tid ≠ tid0, vretRefs(tid0) ≠ null, allocrv(pretRefs, prvHp), allocrv(vretRefs, vrvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, vs), markedp(orl, ps), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), rvownerok(vretRefs, vrvHp, rvowns), unmarkedp(qrl, ps), phead ∈ orl, phead ∈ vs, phead ∈ ps, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ vretRefs(tid0) ∈ vrvHp, ¬ vretRefs(tid0) ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail ∈ orl ∨ vtail ∈ qrl, vhead ∈ orl ∨ vhead ∈ qrl, rvowns(vretRefs(tid0)).pending? ∨ optval00 ≠ Val(rvowns(vretRefs(tid0)).val) ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:rvownerok-def, seq:⊦ rvownerok(vretRefs, vrvHp, rvowns) ↔ (∀ tid. vretRefs(tid) ≠ null → vretRefs(tid) ∈ vrvHp ∧ rvowns(vretRefs(tid)).tid = tid), subst:[vretRefs, vrvHp, rvowns] → [vretRefs, vrvHp, rvowns], <all paths>

Additional information

Used terms: vretRefs, vrvHp, rvowns

Proof step 46 from proof rely-D2-D3 (previous step) (back to tree)

⊦ rvownerok(vretRefs, vrvHp, rvowns) ↔ (∀ tid. vretRefs(tid) ≠ null → vretRefs(tid) ∈ vrvHp ∧ rvowns(vretRefs(tid)).tid = tid)

A spec-lemma

Proof step 47 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, ∅[vretRefs(tid0)] ≠ optval00, orl ≠ [], tid ≠ tid0, vretRefs(tid0) ≠ null, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ vretRefs(tid0) ∈ vrvHp, ¬ vretRefs(tid0) ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), ∀ tid. vretRefs(tid) ≠ null → vretRefs(tid) ∈ vrvHp ∧ rvowns(vretRefs(tid)).tid = tid, rvowns(vretRefs(tid0)).pending? ∨ optval00 ≠ Val(rvowns(vretRefs(tid0)).val), vhead ∈ orl ∨ vhead ∈ qrl, vtail ∈ orl ∨ vtail ∈ qrl, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦

Heuristic quantifier closing applied rule all left

And closed the goal!

Rule argument

pos: left 26 terms: tid0

Additional information

Used terms: tid0

Proof step 48 from proof rely-D2-D3 (previous step) (back to tree)

dom vrvHp = dom prvHp, orl ≠ [], rvref ≠ null, tid ≠ tid0, allocrv(vretRefs, vrvHp), allocrv(pretRefs, prvHp), listp(orl + qrl, vs, ps, owns), markedp(orl, ps), markedp(orl, vs), nextnottid(tid, vhead, vs), noid(tid, orl + qrl, vhead, vs), nonullRetRefs(orl + qrl, vs, vretRefs, pretRefs), noowns(tid, owns), rvownerok(vretRefs, vrvHp, rvowns), rvownerok(pretRefs, prvHp, rvowns), unmarkedp(qrl, ps), phead ∈ ps, phead ∈ vs, phead ∈ orl, ¬ dups(orl + qrl), ¬ dups(orl), ¬ dups(qrl), ¬ prvHpRely(vrvHp, prvHp, prvHp[rvref, optval00]), ¬ rvref ∈ vrvHp, ¬ rvref ∈ prvHp, vs[vhead].next ≠ null ∧ vs[vs[vhead].next].deqID ≠ None → rvowns(vretRefs(vs[vs[vhead].next].deqID.tid)) = Claimed(vs[vs[vhead].next].deqID.tid, vs[vs[vhead].next].val), vhead ∈ orl ∨ vhead ∈ qrl, vtail ∈ orl ∨ vtail ∈ qrl, orl.last = vhead ∧ (qrl ≠ [] → unmarkedp(qrl.tail, vs)) ∨ orl.butlast ≠ [] ∧ orl.butlast.last = vhead ∧ phead ≠ orl.last ∧ unmarkedp(qrl, vs), vtail = (orl + qrl).last ∨ qrl ≠ [] ∧ vtail = (orl + qrl.butlast).last ∧ vs[qrl.last].deqID = None ∧ vs[qrl.last].next = null ⊦

User applied rule apply rewrite lemma

And obtained the following new goals

Rule argument

spec: DMSQ, name:prvHpRely-def, seq:⊦ prvHpRely(vrvHp, prvHp, prvHp0) ↔ (∀ rvref. rvref ∈ prvHp → prvHp0[rvref] = prvHp[rvref] ∨ prvHp0[rvref] = vrvHp[rvref]), subst:[vrvHp, prvHp, prvHp0] → [vrvHp, prvHp, prvHp[rvref, optval00]], <all paths>

Additional information

Used simplifier rules: ⊦ b ∈ st ∧ ¬ a ∈ st → st[a, d][b] = st[b]; (s) Used terms: vrvHp, prvHp, prvHp[rvref, optval00]

Proof step 49 from proof rely-D2-D3 (previous step) (back to tree)

⊦ prvHpRely(vrvHp, prvHp, prvHp0) ↔ (∀ rvref. rvref ∈ prvHp → prvHp0[rvref] = prvHp[rvref] ∨ prvHp0[rvref] = vrvHp[rvref])

A spec-lemma

Proof Tree for rely-D2-D3 (View conclusion)

The following image shows the proof tree of sequent calculus. The proved goal is at the bottom. You can click on the nodes to view additional rules. See here for details on colors in a proof tree.

the proof tree