Commit d504bed6 authored by Michael Neuling's avatar Michael Neuling Committed by Benjamin Herrenschmidt
Browse files

powerpc/kexec: Speedup kexec hash PTE tear down 



Currently for kexec the PTE tear down on 1TB segment systems normally
requires 3 hcalls for each PTE removal. On a machine with 32GB of
memory it can take around a minute to remove all the PTEs.

This optimises the path so that we only remove PTEs that are valid.
It also uses the read 4 PTEs at once HCALL.  For the common case where
a PTEs is invalid in a 1TB segment, this turns the 3 HCALLs per PTE
down to 1 HCALL per 4 PTEs.

This gives an > 10x speedup in kexec times on PHYP, taking a 32GB
machine from around 1 minute down to a few seconds.

Signed-off-by: default avatarMichael Neuling <mikey@neuling.org>
Signed-off-by: default avatarBenjamin Herrenschmidt <benh@kernel.crashing.org>
parent f90ece28

arch/powerpc/platforms/pseries/lpar.c

+20 −13
Original line number Original line Diff line number Diff line
@@ -367,21 +367,28 @@ static void pSeries_lpar_hptab_clear(void) 
{

{

	unsigned long size_bytes = 1UL << ppc64_pft_size;

	unsigned long size_bytes = 1UL << ppc64_pft_size;

	unsigned long hpte_count = size_bytes >> 4;

	unsigned long hpte_count = size_bytes >> 4;

	unsigned long dummy1, dummy2, dword0;

	struct {

		unsigned long pteh;

		unsigned long ptel;

	} ptes[4];

	long lpar_rc;

	long lpar_rc;

	int i;

	int i, j;





	/* TODO: Use bulk call */

	/* Read in batches of 4,

	for (i = 0; i < hpte_count; i++) {

	 * invalidate only valid entries not in the VRMA

		/* dont remove HPTEs with VRMA mappings */

	 * hpte_count will be a multiple of 4

		lpar_rc = plpar_pte_remove_raw(H_ANDCOND, i, HPTE_V_1TB_SEG,

         */

						&dummy1, &dummy2);

	for (i = 0; i < hpte_count; i += 4) {

		if (lpar_rc == H_NOT_FOUND) {

		lpar_rc = plpar_pte_read_4_raw(0, i, (void *)ptes);

			lpar_rc = plpar_pte_read_raw(0, i, &dword0, &dummy1);

		if (lpar_rc != H_SUCCESS)

			if (!lpar_rc && ((dword0 & HPTE_V_VRMA_MASK)

			continue;

				!= HPTE_V_VRMA_MASK))

		for (j = 0; j < 4; j++){

				/* Can be hpte for 1TB Seg. So remove it */

			if ((ptes[j].pteh & HPTE_V_VRMA_MASK) ==

				plpar_pte_remove_raw(0, i, 0, &dummy1, &dummy2);

				HPTE_V_VRMA_MASK)

				continue;

			if (ptes[j].pteh & HPTE_V_VALID)

				plpar_pte_remove_raw(0, i + j, 0,

					&(ptes[j].pteh), &(ptes[j].ptel));

		}

		}

	}

	}

}

}

CRA Git | Maintained and supported by SUSTech CRA and CCSE