lkml.org 
[lkml]   [2011]   [May]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[patch 21/38] firewire: Fix for broken configrom updates in quick succession
2.6.38-stable review patch.  If anyone has any objections, please let us know.

------------------

From: B.J. Buchalter <bj@mhlabs.com>

commit 2e053a27d9d5ad5e0831e002cbf8043836fb2060 upstream.

Current implementation of ohci_set_config_rom() uses a deferred
bus reset via fw_schedule_bus_reset(). If clients add multiple
unit descriptors to the config_rom in quick succession, the
deferred bus reset may not have fired before succeeding update
requests have come in. This can lead to an incorrect partial
update of the config_rom for both addition and removal of
config_rom descriptors, as the ohci_set_config_rom() routine
will return -EBUSY if a previous pending update has not been
completed yet; the requested update just gets dropped on the floor.

This patch recognizes that the "in-flight" update can be modified
until it has been processed by the bus-reset, and the locking
in the bus_reset_tasklet ensures that the update is done atomically
with respect to modifications made by ohci_set_config_rom(). The
-EBUSY error case is simply removed.

[Stefan R: The bug always existed at least theoretically. But it
became easy to trigger since 2.6.36 commit 02d37bed188c "firewire: core:
integrate software-forced bus resets with bus management" which
introduced long mandatory delays between janitorial bus resets.]

Signed-off-by: Benjamin Buchalter <bj@mhlabs.com>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

---
drivers/firewire/ohci.c | 39 +++++++++++++++++++++++++--------------
1 file changed, 25 insertions(+), 14 deletions(-)

--- a/drivers/firewire/ohci.c
+++ b/drivers/firewire/ohci.c
@@ -2163,7 +2163,6 @@ static int ohci_set_config_rom(struct fw
{
struct fw_ohci *ohci;
unsigned long flags;
- int ret = -EBUSY;
__be32 *next_config_rom;
dma_addr_t uninitialized_var(next_config_rom_bus);

@@ -2204,22 +2203,37 @@ static int ohci_set_config_rom(struct fw

spin_lock_irqsave(&ohci->lock, flags);

+ /*
+ * If there is not an already pending config_rom update,
+ * push our new allocation into the ohci->next_config_rom
+ * and then mark the local variable as null so that we
+ * won't deallocate the new buffer.
+ *
+ * OTOH, if there is a pending config_rom update, just
+ * use that buffer with the new config_rom data, and
+ * let this routine free the unused DMA allocation.
+ */
+
if (ohci->next_config_rom == NULL) {
ohci->next_config_rom = next_config_rom;
ohci->next_config_rom_bus = next_config_rom_bus;
+ next_config_rom = NULL;
+ }

- copy_config_rom(ohci->next_config_rom, config_rom, length);
+ copy_config_rom(ohci->next_config_rom, config_rom, length);

- ohci->next_header = config_rom[0];
- ohci->next_config_rom[0] = 0;
+ ohci->next_header = config_rom[0];
+ ohci->next_config_rom[0] = 0;

- reg_write(ohci, OHCI1394_ConfigROMmap,
- ohci->next_config_rom_bus);
- ret = 0;
- }
+ reg_write(ohci, OHCI1394_ConfigROMmap, ohci->next_config_rom_bus);

spin_unlock_irqrestore(&ohci->lock, flags);

+ /* If we didn't use the DMA allocation, delete it. */
+ if (next_config_rom != NULL)
+ dma_free_coherent(ohci->card.device, CONFIG_ROM_SIZE,
+ next_config_rom, next_config_rom_bus);
+
/*
* Now initiate a bus reset to have the changes take
* effect. We clean up the old config rom memory and DMA
@@ -2227,13 +2241,10 @@ static int ohci_set_config_rom(struct fw
* controller could need to access it before the bus reset
* takes effect.
*/
- if (ret == 0)
- fw_schedule_bus_reset(&ohci->card, true, true);
- else
- dma_free_coherent(ohci->card.device, CONFIG_ROM_SIZE,
- next_config_rom, next_config_rom_bus);

- return ret;
+ fw_schedule_bus_reset(&ohci->card, true, true);
+
+ return 0;
}

static void ohci_send_request(struct fw_card *card, struct fw_packet *packet)



\
 
 \ /
  Last update: 2011-05-06 02:19    [W:0.427 / U:0.208 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site