ide: remove hardcoded 2GiB transactional limit

Not that you can request a >2GiB transaction, but that's why checking
for it makes no sense anymore.

With the newer 'limit' parameter to prepare_buf, we no longer need a
static limit. The maximum limit is still 2GiB, but the limit parameter
is set to the current transaction size, which cannot surpass 32MiB
(512 * 65536). If the PRDT surpasses the transactional size, then,
we'll just carry out the normative underflow handling pathways instead
of needing an extra, strange pathway that worries about hitting some
logistical cap for the largest sglist we can support -- we'll never
even attempt to build one that big anymore.

Reported-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1445902682-20051-1-git-send-email-jsnow@redhat.com

1 files changed, 0 insertions, 7 deletions

diff --git a/hw/ide/pci.c b/hw/ide/pci.c
index d31ff885b7..9c54b378d6 100644
--- a/hw/ide/pci.c
+++ b/hw/ide/pci.c
@@ -103,13 +103,6 @@ static int32_t bmdma_prepare_buf(IDEDMA *dma, int32_t limit)
                 qemu_sglist_add(&s->sg, bm->cur_prd_addr, sg_len);
             }
 
-            /* Note: We limit the max transfer to be 2GiB.
-             * This should accommodate the largest ATA transaction
-             * for LBA48 (65,536 sectors) and 32K sector sizes. */
-            if (s->sg.size > INT32_MAX) {
-                error_report("IDE: sglist describes more than 2GiB.");
-                break;
-            }
             bm->cur_prd_addr += l;
             bm->cur_prd_len -= l;
             s->io_buffer_size += l;