#include "bits/stdc++.h"
using namespace std;
typedef int ll;
typedef pair<ll, ll> pll;
typedef vector<ll> vll;
#pragma GCC optimize("O3")
#pragma GCC target("avx2")
const ll N = 305;
ll n, r, c, x[N], y[N], ans = 2e9, _x[N], _y[N], ps[N], pL[N][N], pR[N][N], pmx[N][N],
itl[N], itr[N], fL[N * N], fR[N * N], fmx[N * N], cL[N * N], cR[N * N], cmx[N * N];
void prec() {
for (ll gl = 0; gl < n; gl++) {
for (ll gr = gl; gr < n; gr++) {
vector<ll> a;
for (ll i = gl; i <= gr; i++) a.push_back(y[i]);
sort(a.begin(), a.end());
pL[gl][gr] = a[0];
pR[gl][gr] = c - 1 - a[gr - gl];
for (ll i = 1; i < a.size(); i++) pmx[gl][gr] = max(pmx[gl][gr], a[i] - a[i - 1] - 1);
fL[gl * N + gr] = pL[gl][gr];
fR[gl * N + gr] = pR[gl][gr];
fmx[gl * N + gr] = pmx[gl][gr];
}
}
sort(fL, fL + N * N);
sort(fR, fR + N * N);
sort(fmx, fmx + N * N);
for (ll gl = 0; gl < n; gl++) {
for (ll gr = gl; gr < n; gr++) {
pL[gl][gr] = lower_bound(fL, fL + N * N, pL[gl][gr]) - fL;
pR[gl][gr] = lower_bound(fR, fR + N * N, pR[gl][gr]) - fR;
pmx[gl][gr] = lower_bound(fmx, fmx + N * N, pmx[gl][gr]) - fmx;
}
}
}
signed main() {
ios::sync_with_stdio(0), cin.tie(0), cout.tie(0);
cin >> r >> c >> n;
for (ll i = 0; i < n; i++) cin >> x[i] >> y[i], x[i]--, y[i]--;
iota(ps, ps + n, 0);
sort(ps, ps + n, [&](ll a, ll b) { return x[a] < x[b]; });
for (ll i = 0; i < n; i++) _x[i] = x[ps[i]], _y[i] = y[ps[i]];
for (ll i = 0; i < n; i++) swap(x[i], _x[i]), swap(y[i], _y[i]);
prec();
for (ll up = 0; up < n; up++) { // n
vll cup;
for (ll i = 0; i < n; i++) {
cup.push_back(r - 1 - x[i]);
for (ll j = i; j < n; j++) {
cup.push_back(max(0, abs(x[i] - x[j]) - 1 - x[up]));
}
}
memset(cL, 0, sizeof cL);
memset(cR, 0, sizeof cR);
memset(cmx, 0, sizeof cmx);
sort(cup.begin(), cup.end());
cup.resize(unique(cup.begin(), cup.end()) - cup.begin());
ll il = 0, ir = -1;
vector<array<ll, 2>> s;
ll iL = -1, iR = -1, imx = -1;
ll ce = 0;
for (ll i = 0; i < n; i++) { // n^3
if (x[i] + cup[0] + 1 < r) {
while (il < n && x[il] + cup[0] < x[i] + cup[0] + 1) il++;
while (ir + 1 < n && x[ir + 1] - x[up] <= x[i] + cup[0] + 1) ir++;
itl[i] = il;
itr[i] = ir;
if (ir < il) ce++;
else {
iL = max(iL, pL[itl[i]][itr[i]]);
iR = max(iR, pR[itl[i]][itr[i]]);
imx = max(imx, pmx[itl[i]][itr[i]]);
cL[pL[itl[i]][itr[i]]]++;
cR[pR[itl[i]][itr[i]]]++;
cmx[pmx[itl[i]][itr[i]]]++;
// sL.insert(pL[il][ir]);
// sR.insert(pR[il][ir]);
// smx.insert(pmx[il][ir]);
}
ll pr = ir;
while (pr + 1 < n) {
pr++;
s.push_back({(x[pr] - x[up]) - (x[i] + cup[0] + 1), i});
}
s.push_back({r - (x[i] + cup[0] + 1), i});
}
}
il = 0, ir = -1;
while (il < n && x[il] + cup[0] < 0) il++;
while (ir + 1 < n && x[ir + 1] - x[up] <= 0) ir++;
sort(s.begin(), s.end());
ll it = 0;
ans = min(ans, (ce ? (ll) 2e9 : max(max(fmx[pmx[il][ir]], (imx != -1 ? fmx[imx] : 0)),
max(fL[pL[il][ir]], (iL != -1 ? fL[iL] : 0)) +
max(fR[pR[il][ir]], (iR != -1 ? fR[iR] : 0)))) +
cup[0] + x[up]);
for (ll dn = 1; dn < cup.size(); dn++) {
while (it < s.size() && s[it][0] <= cup[dn]) {
ll i = s[it][1];
if (itr[i] < itl[i]) ce--;
else {
cL[pL[itl[i]][itr[i]]]--;
cR[pR[itl[i]][itr[i]]]--;
cmx[pmx[itl[i]][itr[i]]]--;
while (iL >= 0 && !cL[iL]) iL--;
while (iR >= 0 && !cR[iR]) iR--;
while (imx >= 0 && !cmx[imx]) imx--;
}
itr[i]++;
if (itr[i] != n) {
iL = max(iL, pL[itl[i]][itr[i]]);
iR = max(iR, pR[itl[i]][itr[i]]);
imx = max(imx, pmx[itl[i]][itr[i]]);
cL[pL[itl[i]][itr[i]]]++;
cR[pR[itl[i]][itr[i]]]++;
cmx[pmx[itl[i]][itr[i]]]++;
}
it++;
}
while (il < n && x[il] + cup[0] < 0) il++;
while (ir + 1 < n && x[ir + 1] - x[up] <= 0) ir++;
ans = min(ans, (ce ? (ll) 2e9 : max(max(fmx[pmx[il][ir]], (imx != -1 ? fmx[imx] : 0)),
max(fL[pL[il][ir]], (iL != -1 ? fL[iL] : 0)) +
max(fR[pR[il][ir]], (iR != -1 ? fR[iR] : 0)))) +
cup[dn] + x[up]);
}
}
cout << ans;
}
