From: Yi Zou Date: Wed, 21 Oct 2009 23:27:58 +0000 (-0700) Subject: [SCSI] libfc: Fix frags in frame exceeding SKB_MAX_FRAGS in fc_fcp_send_data X-Git-Tag: v2.6.33-rc1~344^2~149 X-Git-Url: https://bbs.cooldavid.org/git/?a=commitdiff_plain;h=d37322a43ebac79eef417149f5696390cf8872db;p=net-next-2.6.git [SCSI] libfc: Fix frags in frame exceeding SKB_MAX_FRAGS in fc_fcp_send_data In case of sequence offload, in fc_fcp_send_data(), the skb_fill_page_info() called may end up adding more frags to the skb_shinfo(fp_skb(fp))->frags[], exceeding SKB_MAX_FRAGS, this eventually corrupts the memory. I am adding the FR_FRAME_SG_LEN back, but as SKB_MAX_FRAGS -1, leaving 1 for our fcoe_eof_crc page. And send will be broken into multiple large sends if the frame already contains more frags than skb handle. Signed-off-by: Yi Zou Signed-off-by: Robert Love Signed-off-by: James Bottomley --- diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c index 40ed7442d9d..28bfe1c2c50 100644 --- a/drivers/scsi/libfc/fc_fcp.c +++ b/drivers/scsi/libfc/fc_fcp.c @@ -574,7 +574,8 @@ static int fc_fcp_send_data(struct fc_fcp_pkt *fsp, struct fc_seq *seq, tlen -= sg_bytes; remaining -= sg_bytes; - if (tlen) + if ((skb_shinfo(fp_skb(fp))->nr_frags < FC_FRAME_SG_LEN) && + (tlen)) continue; /* diff --git a/include/scsi/fc_frame.h b/include/scsi/fc_frame.h index c35d2383cc2..148126dcf9e 100644 --- a/include/scsi/fc_frame.h +++ b/include/scsi/fc_frame.h @@ -37,6 +37,9 @@ #define FC_FRAME_HEADROOM 32 /* headroom for VLAN + FCoE headers */ #define FC_FRAME_TAILROOM 8 /* trailer space for FCoE */ +/* Max number of skb frags allowed, reserving one for fcoe_crc_eof page */ +#define FC_FRAME_SG_LEN (MAX_SKB_FRAGS - 1) + #define fp_skb(fp) (&((fp)->skb)) #define fr_hdr(fp) ((fp)->skb.data) #define fr_len(fp) ((fp)->skb.len)