From: Leonardo Bras <leobras.c@gmail.com>
To: Michael Ellerman <mpe@ellerman.id.au>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Paul Mackerras <paulus@samba.org>,
Leonardo Bras <leobras.c@gmail.com>,
Joel Stanley <joel@jms.id.au>,
Christophe Leroy <christophe.leroy@c-s.fr>,
Alexey Kardashevskiy <aik@ozlabs.ru>,
Thiago Jung Bauermann <bauerman@linux.ibm.com>,
Ram Pai <linuxram@us.ibm.com>,
Brian King <brking@linux.vnet.ibm.com>,
Murilo Fossa Vicentini <muvic@linux.ibm.com>,
David Dai <zdai@linux.vnet.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org
Subject: [PATCH v2 03/14] powerpc/kernel/iommu: Align size for IOMMU_PAGE_SIZE() to save TCEs
Date: Fri, 11 Sep 2020 14:07:28 -0300 [thread overview]
Message-ID: <20200911170738.82818-4-leobras.c@gmail.com> (raw)
In-Reply-To: <20200911170738.82818-1-leobras.c@gmail.com>
Cc: linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org,
Currently both iommu_alloc_coherent() and iommu_free_coherent() align the
desired allocation size to PAGE_SIZE, and gets system pages and IOMMU
mappings (TCEs) for that value.
When IOMMU_PAGE_SIZE < PAGE_SIZE, this behavior may cause unnecessary
TCEs to be created for mapping the whole system page.
Example:
- PAGE_SIZE = 64k, IOMMU_PAGE_SIZE() = 4k
- iommu_alloc_coherent() is called for 128 bytes
- 1 system page (64k) is allocated
- 16 IOMMU pages (16 x 4k) are allocated (16 TCEs used)
It would be enough to use a single TCE for this, so 15 TCEs are
wasted in the process.
Update iommu_*_coherent() to make sure the size alignment happens only
for IOMMU_PAGE_SIZE() before calling iommu_alloc() and iommu_free().
Also, on iommu_range_alloc(), replace ALIGN(n, 1 << tbl->it_page_shift)
with IOMMU_PAGE_ALIGN(n, tbl), which is easier to read and does the
same.
Signed-off-by: Leonardo Bras <leobras.c@gmail.com>
---
arch/powerpc/kernel/iommu.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/arch/powerpc/kernel/iommu.c b/arch/powerpc/kernel/iommu.c
index 9704f3f76e63..7961645a6980 100644
--- a/arch/powerpc/kernel/iommu.c
+++ b/arch/powerpc/kernel/iommu.c
@@ -237,10 +237,9 @@ static unsigned long iommu_range_alloc(struct device *dev,
}
if (dev)
- boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1,
- 1 << tbl->it_page_shift);
+ boundary_size = IOMMU_PAGE_ALIGN(dma_get_seg_boundary(dev) + 1, tbl);
else
- boundary_size = ALIGN(1UL << 32, 1 << tbl->it_page_shift);
+ boundary_size = IOMMU_PAGE_ALIGN(1UL << 32, tbl);
/* 4GB boundary for iseries_hv_alloc and iseries_hv_map */
n = iommu_area_alloc(tbl->it_map, limit, start, npages, tbl->it_offset,
@@ -858,6 +857,7 @@ void *iommu_alloc_coherent(struct device *dev, struct iommu_table *tbl,
unsigned int order;
unsigned int nio_pages, io_order;
struct page *page;
+ size_t size_io = size;
size = PAGE_ALIGN(size);
order = get_order(size);
@@ -884,8 +884,9 @@ void *iommu_alloc_coherent(struct device *dev, struct iommu_table *tbl,
memset(ret, 0, size);
/* Set up tces to cover the allocated range */
- nio_pages = size >> tbl->it_page_shift;
- io_order = get_iommu_order(size, tbl);
+ size_io = IOMMU_PAGE_ALIGN(size_io, tbl);
+ nio_pages = size_io >> tbl->it_page_shift;
+ io_order = get_iommu_order(size_io, tbl);
mapping = iommu_alloc(dev, tbl, ret, nio_pages, DMA_BIDIRECTIONAL,
mask >> tbl->it_page_shift, io_order, 0);
if (mapping == DMA_MAPPING_ERROR) {
@@ -900,10 +901,9 @@ void iommu_free_coherent(struct iommu_table *tbl, size_t size,
void *vaddr, dma_addr_t dma_handle)
{
if (tbl) {
- unsigned int nio_pages;
+ size_t size_io = IOMMU_PAGE_ALIGN(size, tbl);
+ unsigned int nio_pages = size_io >> tbl->it_page_shift;
- size = PAGE_ALIGN(size);
- nio_pages = size >> tbl->it_page_shift;
iommu_free(tbl, dma_handle, nio_pages);
size = PAGE_ALIGN(size);
free_pages((unsigned long)vaddr, get_order(size));
--
2.25.4
next prev parent reply other threads:[~2020-09-11 17:21 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-11 17:07 [PATCH v2 00/14] DDW Indirect Mapping Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 01/14] powerpc/pseries/iommu: Replace hard-coded page shift Leonardo Bras
2020-09-29 3:56 ` Alexey Kardashevskiy
2020-09-29 18:25 ` Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 02/14] powerpc/pseries/iommu: Makes sure IOMMU_PAGE_SIZE <= PAGE_SIZE Leonardo Bras
2020-09-29 3:57 ` Alexey Kardashevskiy
2020-09-11 17:07 ` Leonardo Bras [this message]
2020-09-29 3:57 ` [PATCH v2 03/14] powerpc/kernel/iommu: Align size for IOMMU_PAGE_SIZE() to save TCEs Alexey Kardashevskiy
2020-09-11 17:07 ` [PATCH v2 04/14] powerpc/kernel/iommu: Use largepool as a last resort when !largealloc Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 05/14] powerpc/kernel/iommu: Add new iommu_table_in_use() helper Leonardo Bras
2020-09-29 3:57 ` Alexey Kardashevskiy
2021-04-11 6:55 ` Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 06/14] powerpc/pseries/iommu: Add iommu_pseries_alloc_table() helper Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 07/14] powerpc/pseries/iommu: Add ddw_list_new_entry() helper Leonardo Bras
2020-09-29 3:57 ` Alexey Kardashevskiy
2020-09-11 17:07 ` [PATCH v2 08/14] powerpc/pseries/iommu: Allow DDW windows starting at 0x00 Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 09/14] powerpc/pseries/iommu: Add ddw_property_create() and refactor enable_ddw() Leonardo Bras
2020-09-29 3:56 ` Alexey Kardashevskiy
2021-04-11 7:52 ` Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 10/14] powerpc/pseries/iommu: Reorganize iommu_table_setparms*() with new helper Leonardo Bras
2020-09-29 3:56 ` Alexey Kardashevskiy
2021-04-11 8:16 ` Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 11/14] powerpc/pseries/iommu: Update remove_dma_window() to accept property name Leonardo Bras
2020-09-29 3:56 ` Alexey Kardashevskiy
2021-04-13 5:44 ` Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 12/14] powerpc/pseries/iommu: Find existing DDW with given " Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 13/14] powerpc/pseries/iommu: Make use of DDW for indirect mapping Leonardo Bras
2020-09-29 3:56 ` Alexey Kardashevskiy
2021-04-13 5:49 ` Leonardo Bras
2021-04-13 7:18 ` Alexey Kardashevskiy
2021-04-13 7:33 ` Leonardo Bras
2021-04-13 7:41 ` Alexey Kardashevskiy
2021-04-13 7:58 ` Leonardo Bras
2021-04-13 8:24 ` Alexey Kardashevskiy
2021-04-13 23:01 ` Leonardo Bras
2020-09-11 17:07 ` [PATCH v2 14/14] powerpc/pseries/iommu: Rename "direct window" to "dma window" Leonardo Bras
2020-09-29 3:55 ` Alexey Kardashevskiy
2020-09-29 20:54 ` Leonardo Bras
2020-09-30 7:29 ` Alexey Kardashevskiy
2021-04-13 6:03 ` Leonardo Bras
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200911170738.82818-4-leobras.c@gmail.com \
--to=leobras.c@gmail.com \
--cc=aik@ozlabs.ru \
--cc=bauerman@linux.ibm.com \
--cc=benh@kernel.crashing.org \
--cc=brking@linux.vnet.ibm.com \
--cc=christophe.leroy@c-s.fr \
--cc=joel@jms.id.au \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=linuxram@us.ibm.com \
--cc=mpe@ellerman.id.au \
--cc=muvic@linux.ibm.com \
--cc=paulus@samba.org \
--cc=zdai@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).