Hi Gustavo.

Sorry for my late reply.

Yes, I will remove this patch and replaces it with your proposed patch[1] from 
the V9 patch email thread.

I see that Bryan has already tested your new patch and it works.  I agree the 
new patch is a much simpler/better version.

[1] 
https://lore.kernel.org/linux-nvme/[email protected]/#t

I will add this to my V11 patch series.

/John

On 9/26/25 5:00 AM, Gustavo A. R. Silva wrote:
Hi,

Shouldn't this patch be removed from this series, since it's going to be
reverted anyways?

Thanks
-Gustavo

On 9/26/25 02:02, John Meneghini wrote:
From: "Gustavo A. R. Silva" <[email protected]>

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.

   >  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`.

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`. This is wrapped using the
   `TRAILING_OVERLAP()` macro helper, which effectively creates a union
   between flexible-array member `default_item.iocb` and
   `__default_item_iocb`.

   Since `struct pure_item` now contains a flexible-array member, the
   helper must be placed at the end of `struct scsi_qla_host` to prevent
   a `-Wflex-array-member-not-at-end` warning.

   `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 between `size` and `iocb`. The resulting
size is the same, but iocb is shifted 2-bytes (the original `purex_item`
structure was padded at the end, after the 64-byte defined array size).
I don't think this is a problem.

In qla_os.c:qla24xx_process_purex_rdp()

To avoid a null pointer dereference the vha->default_item should be set
to 0 last if the item pointer passed to the function matches.  Also use
a local variable to avoid multiple de-referencing of the item.

Tested-by: Bryan Gurney <[email protected]>
Co-developed-by: Chris Leech <[email protected]>
Signed-off-by: Chris Leech <[email protected]>
Signed-off-by: Gustavo A. R. Silva <[email protected]>
---
  drivers/scsi/qla2xxx/qla_def.h  | 10 ++++++----
  drivers/scsi/qla2xxx/qla_isr.c  | 17 ++++++++---------
  drivers/scsi/qla2xxx/qla_nvme.c |  2 +-
  drivers/scsi/qla2xxx/qla_os.c   |  9 ++++++---
  4 files changed, 21 insertions(+), 17 deletions(-)

diff --git a/drivers/scsi/qla2xxx/qla_def.h b/drivers/scsi/qla2xxx/qla_def.h
index cb95b7b12051..604e66bead1e 100644
--- a/drivers/scsi/qla2xxx/qla_def.h
+++ b/drivers/scsi/qla2xxx/qla_def.h
@@ -4890,9 +4890,7 @@ struct purex_item {
                   struct purex_item *pkt);
      atomic_t in_use;
      uint16_t size;
-    struct {
-        uint8_t iocb[64];
-    } iocb;
+    uint8_t iocb[] __counted_by(size);
  };
  #include "qla_edif.h"
@@ -5101,7 +5099,6 @@ typedef struct scsi_qla_host {
          struct list_head head;
          spinlock_t lock;
      } purex_list;
-    struct purex_item default_item;
      struct name_list_extended gnl;
      /* Count of active session/fcport */
@@ -5130,6 +5127,11 @@ typedef struct scsi_qla_host {
  #define DPORT_DIAG_IN_PROGRESS                 BIT_0
  #define DPORT_DIAG_CHIP_RESET_IN_PROGRESS      BIT_1
      uint16_t dport_status;
+
+    /* Must be last --ends in a flexible-array member. */
+    TRAILING_OVERLAP(struct purex_item, default_item, iocb,
+        uint8_t __default_item_iocb[QLA_DEFAULT_PAYLOAD_SIZE];
+    );
  } scsi_qla_host_t;
  struct qla27xx_image_status {
diff --git a/drivers/scsi/qla2xxx/qla_isr.c b/drivers/scsi/qla2xxx/qla_isr.c
index 8ff8781dae47..ccb044693dcb 100644
--- a/drivers/scsi/qla2xxx/qla_isr.c
+++ b/drivers/scsi/qla2xxx/qla_isr.c
@@ -1078,17 +1078,17 @@ 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);
      } else {
          if (atomic_inc_return(&vha->default_item.in_use) == 1) {
              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);
          }
      }
      if (!item) {
@@ -1128,17 +1128,16 @@ 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 8ee2e337c9e1..92488890bc04 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 d4b484c0fd9d..32437bae1a30 100644
--- a/drivers/scsi/qla2xxx/qla_os.c
+++ b/drivers/scsi/qla2xxx/qla_os.c
@@ -6459,9 +6459,12 @@ 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
+    scsi_qla_host_t *base_vha = item->vha;
+
+    if (item == &base_vha->default_item) {
+        memset(&base_vha->__default_item_iocb, 0, QLA_DEFAULT_PAYLOAD_SIZE);
+        memset(&base_vha->default_item, 0, sizeof(struct purex_item));
+    } else
          kfree(item);
  }



Reply via email to