From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
To: Chris Leech <cleech@redhat.com>, linux-scsi@vger.kernel.org
Cc: Nilesh Javali <njavali@marvell.com>, Kees Cook <kees@kernel.org>,
Bryan Gurney <bgurney@redhat.com>,
"Gustavo A . R . Silva" <gustavoars@kernel.org>,
John Meneghini <jmeneghi@redhat.com>
Subject: Re: [PATCH v2 1/1] scsi: qla2xxx: replace non-standard flexible array purex_item.iocb
Date: Mon, 28 Jul 2025 13:43:10 -0600 [thread overview]
Message-ID: <a1c61211-816e-4479-81ce-e71a0d2b8ec2@embeddedor.com> (raw)
In-Reply-To: <20250728185725.2501761-1-cleech@redhat.com>
On 28/07/25 12:57, Chris Leech wrote:
> purex_item.iocb is defined as a 64-element u8 array, but 64 is the
> minimum size and it can be allocated larger.
> This makes it a standard empty flex array.
>
> This was motivated by field-spanning write warnings during FPIN testing.
> https://lore.kernel.org/linux-nvme/20250709211919.49100-1-bgurney@redhat.com/
>
> > kernel: memcpy: detected field-spanning write (size 60) of single field
> > "((uint8_t *)fpin_pkt + buffer_copy_offset)"
> > at drivers/scsi/qla2xxx/qla_isr.c:1221 (size 44)
>
> I removed the outer wrapper from the iocb flex array, so that it can be
> linked to purex_item.size with __counted_by.
>
> The non-flex-array portion of purex_item is now defined within the
> struct_group_tagged macro to create a type definition for just the
> header. This was then used for the scsi_qla_host.default_item
> definition to deal with -Wflex-array-member-not-at-end warnings.
>
> These changes remove the default minimum 64-byte allocation, requiring
> further changes.
>
> In struct scsi_qla_host the embedded default_item is now followed by
> __default_item_iocb[QLA_DEFAULT_PAYLOAD_SIZE] to reserve space that
> will be used as default_item.iocb
>
> qla24xx_alloc_purex_item is adjusted to no longer expect the default
> minimum size to be part of sizeof(struct purex_item), the entire
> flexible array size is added to the structure size for allocation.
>
> This also slightly changes the layout of the purex_item struct, as
> 2-bytes of padding are added to purex_item_hdr and therefor between size
> and iocb. The resulting size is the same, but iocb is shifted 2-bytes
> (the old purex_item was padded at the end, after the 64-byte defined
> array size). I don't think this is a problem.
>
> Tested-by: Bryan Gurney <bgurney@redhat.com>
> Signed-off-by: Chris Leech <cleech@redhat.com>
> ---
> drivers/scsi/qla2xxx/qla_def.h | 23 ++++++++++++-----------
> drivers/scsi/qla2xxx/qla_isr.c | 19 +++++++++----------
> drivers/scsi/qla2xxx/qla_nvme.c | 2 +-
> drivers/scsi/qla2xxx/qla_os.c | 9 +++++----
> 4 files changed, 27 insertions(+), 26 deletions(-)
>
> diff --git a/drivers/scsi/qla2xxx/qla_def.h b/drivers/scsi/qla2xxx/qla_def.h
> index cb95b7b12051d..0e7d201b1feb0 100644
> --- a/drivers/scsi/qla2xxx/qla_def.h
> +++ b/drivers/scsi/qla2xxx/qla_def.h
> @@ -4883,16 +4883,16 @@ struct active_regions {
> * is variable) starting at "iocb".
> */
> struct purex_item {
> - void *purls_context;
> - struct list_head list;
> - struct scsi_qla_host *vha;
> - void (*process_item)(struct scsi_qla_host *vha,
> - struct purex_item *pkt);
> - atomic_t in_use;
> - uint16_t size;
> - struct {
> - uint8_t iocb[64];
> - } iocb;
> + struct_group_tagged(purex_item_hdr, __hdr,
> + void *purls_context;
> + struct list_head list;
> + struct scsi_qla_host *vha;
> + void (*process_item)(struct scsi_qla_host *vha,
> + struct purex_item *pkt);
> + atomic_t in_use;
> + uint16_t size;
> + );
> + uint8_t iocb[] __counted_by(size);
> };
>
> #include "qla_edif.h"
> @@ -5101,7 +5101,8 @@ typedef struct scsi_qla_host {
> struct list_head head;
> spinlock_t lock;
> } purex_list;
> - struct purex_item default_item;
> + struct purex_item_hdr default_item;
> + uint8_t __default_item_iocb[QLA_DEFAULT_PAYLOAD_SIZE];
>
> struct name_list_extended gnl;
> /* Count of active session/fcport */
> diff --git a/drivers/scsi/qla2xxx/qla_isr.c b/drivers/scsi/qla2xxx/qla_isr.c
> index fe98c76e9be32..a00c06a9898ec 100644
> --- a/drivers/scsi/qla2xxx/qla_isr.c
> +++ b/drivers/scsi/qla2xxx/qla_isr.c
> @@ -1077,17 +1077,17 @@ static struct purex_item *
> qla24xx_alloc_purex_item(scsi_qla_host_t *vha, uint16_t size)
> {
> struct purex_item *item = NULL;
> - uint8_t item_hdr_size = sizeof(*item);
>
> if (size > QLA_DEFAULT_PAYLOAD_SIZE) {
> - item = kzalloc(item_hdr_size +
> - (size - QLA_DEFAULT_PAYLOAD_SIZE), GFP_ATOMIC);
> + item = kzalloc(struct_size(item, iocb, size), GFP_ATOMIC);
With the inclusion of `counted_by`, I think `item->size` should be updated
here:
item->size = size;
> } else {
> if (atomic_inc_return(&vha->default_item.in_use) == 1) {
> - item = &vha->default_item;
> + item = (struct purex_item *)&vha->default_item;
> goto initialize_purex_header;
> } else {
> - item = kzalloc(item_hdr_size, GFP_ATOMIC);
> + item = kzalloc(
> + struct_size(item, iocb, QLA_DEFAULT_PAYLOAD_SIZE),
> + GFP_ATOMIC);
...and here:
item->size = QLA_DEFAULT_PAYLOAD_SIZE;
Then remove `item->size = size;` just before `return item;`
Thanks
-Gustavo
> }
> }
> if (!item) {
> @@ -1127,17 +1127,16 @@ qla24xx_queue_purex_item(scsi_qla_host_t *vha, struct purex_item *pkt,
> * @vha: SCSI driver HA context
> * @pkt: ELS packet
> */
> -static struct purex_item
> -*qla24xx_copy_std_pkt(struct scsi_qla_host *vha, void *pkt)
> +static struct purex_item *
> +qla24xx_copy_std_pkt(struct scsi_qla_host *vha, void *pkt)
> {
> struct purex_item *item;
>
> - item = qla24xx_alloc_purex_item(vha,
> - QLA_DEFAULT_PAYLOAD_SIZE);
> + item = qla24xx_alloc_purex_item(vha, QLA_DEFAULT_PAYLOAD_SIZE);
> if (!item)
> return item;
>
> - memcpy(&item->iocb, pkt, sizeof(item->iocb));
> + memcpy(&item->iocb, pkt, QLA_DEFAULT_PAYLOAD_SIZE);
> return item;
> }
>
> diff --git a/drivers/scsi/qla2xxx/qla_nvme.c b/drivers/scsi/qla2xxx/qla_nvme.c
> index 8ee2e337c9e1b..92488890bc04e 100644
> --- a/drivers/scsi/qla2xxx/qla_nvme.c
> +++ b/drivers/scsi/qla2xxx/qla_nvme.c
> @@ -1308,7 +1308,7 @@ void qla2xxx_process_purls_iocb(void **pkt, struct rsp_que **rsp)
>
> ql_dbg(ql_dbg_unsol, vha, 0x2121,
> "PURLS OP[%01x] size %d xchg addr 0x%x portid %06x\n",
> - item->iocb.iocb[3], item->size, uctx->exchange_address,
> + item->iocb[3], item->size, uctx->exchange_address,
> fcport->d_id.b24);
> /* +48 0 1 2 3 4 5 6 7 8 9 A B C D E F
> * ----- -----------------------------------------------
> diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
> index d4b484c0fd9d7..ef8a4fce0e03c 100644
> --- a/drivers/scsi/qla2xxx/qla_os.c
> +++ b/drivers/scsi/qla2xxx/qla_os.c
> @@ -3971,7 +3971,7 @@ qla2x00_remove_one(struct pci_dev *pdev)
> static inline void
> qla24xx_free_purex_list(struct purex_list *list)
> {
> - struct purex_item *item, *next;
> + struct purex_item_hdr *item, *next;
> ulong flags;
>
> spin_lock_irqsave(&list->lock, flags);
> @@ -6459,9 +6459,10 @@ void qla24xx_process_purex_rdp(struct scsi_qla_host *vha,
> void
> qla24xx_free_purex_item(struct purex_item *item)
> {
> - if (item == &item->vha->default_item)
> - memset(&item->vha->default_item, 0, sizeof(struct purex_item));
> - else
> + if (item == (struct purex_item *)&item->vha->default_item) {
> + memset(&item->vha->default_item, 0, sizeof(struct purex_item_hdr));
> + memset(&item->vha->__default_item_iocb, 0, QLA_DEFAULT_PAYLOAD_SIZE);
> + } else
> kfree(item);
> }
>
next prev parent reply other threads:[~2025-07-28 19:43 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-07-25 21:27 [PATCH 0/2] scsi: qla2xxx: flexible array / field-spanning write issue Chris Leech
2025-07-25 21:27 ` [PATCH 1/2] scsi: qla2xxx: replace non-standard flexible array purex_item.iocb Chris Leech
2025-07-25 21:54 ` Kees Cook
2025-07-28 18:57 ` [PATCH v2 1/1] " Chris Leech
2025-07-28 19:43 ` Gustavo A. R. Silva [this message]
2025-07-28 19:54 ` Gustavo A. R. Silva
2025-07-28 21:15 ` Chris Leech
2025-07-28 22:55 ` Gustavo A. R. Silva
2025-07-28 23:52 ` Chris Leech
2025-07-29 1:37 ` Gustavo A. R. Silva
2025-07-29 2:20 ` Gustavo A. R. Silva
2025-07-30 0:04 ` Chris Leech
2025-07-30 1:33 ` Gustavo A. R. Silva
2025-07-30 3:44 ` Gustavo A. R. Silva
2025-07-30 15:43 ` Bryan Gurney
2025-07-30 20:55 ` Gustavo A. R. Silva
2025-07-25 21:27 ` [PATCH 2/2] scsi: qla2xxx: unwrap purex_item.iocb.iocb so that __counted_by can be used Chris Leech
2025-07-25 21:56 ` Kees Cook
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=a1c61211-816e-4479-81ce-e71a0d2b8ec2@embeddedor.com \
--to=gustavo@embeddedor.com \
--cc=bgurney@redhat.com \
--cc=cleech@redhat.com \
--cc=gustavoars@kernel.org \
--cc=jmeneghi@redhat.com \
--cc=kees@kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=njavali@marvell.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