| Message ID | 6EA3E0BCC03CC34B89B01BD57ECBC718B17AFE41@POBOX.postoffice.danego.net (mailing list archive) |
|---|---|
| State | Rejected |
| Delegated to: | Kalle Valo |
| Headers | show |
I have found the original post of this patch: http://lists.infradead.org/pipermail/linux-arm-kernel/2013-November/210624.html So all credits go to the original author, Jason Liu. After reading Kalle's comments, I have isolated the cause of my problem and it turns out to be the major bug in drivers/net/wireless/ath/ath6kl/hif.h, causing hif_scatter_item scat_list to corrupt scat_q_depth. At bit unfortunate that this fix wasn't taken in... Even though I am using GCC 4.8.1, I'm probably not stressing the target hard enough to require the work-around in drivers/net/wireless/ath/ath6kl/sdio.c -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi Robert, On Fri, Nov 6, 2015 at 7:42 PM, Robert Deliën <robert@delien.nl> wrote: > This patch fixes broken hif-scatter in ath6kl SDIO. It is not my work; I have found it as a difference between two Digi distributed Kernel source trees. No commit comment was recorded. I am sure there are better/nicer/cleaner way, but copying it as I found it is yet another difference less in the future. Feel free to use it in any way you see fit. > > Signed-off-by: Robert Delien <robert@delien.nl> Did you ever re-send this patch? It looks like a fairly important fix and was never applied. Thanks, Julian Calaby > > > Index: drivers/net/wireless/ath/ath6kl/sdio.c > =================================================================== > --- drivers/net/wireless/ath/ath6kl/sdio.c (revision 715) > +++ drivers/net/wireless/ath/ath6kl/sdio.c (revision 716) > @@ -222,6 +222,7 @@ > struct mmc_data *data) > { > struct scatterlist *sg; > + struct hif_scatter_item *scat_list; > int i; > > data->blksz = HIF_MBOX_BLOCK_SIZE; > @@ -240,14 +241,14 @@ > sg = scat_req->sgentries; > sg_init_table(sg, scat_req->scat_entries); > > + scat_list = &scat_req->scat_list[0]; > + > /* assemble SG list */ > - for (i = 0; i < scat_req->scat_entries; i++, sg++) { > + for (i = 0; i < scat_req->scat_entries; i++, sg++, scat_list++) { > ath6kl_dbg(ATH6KL_DBG_SCATTER, "%d: addr:0x%p, len:%d\n", > - i, scat_req->scat_list[i].buf, > - scat_req->scat_list[i].len); > + i, scat_list->buf, scat_list->len); > > - sg_set_buf(sg, scat_req->scat_list[i].buf, > - scat_req->scat_list[i].len); > + sg_set_buf(sg, scat_list->buf, scat_list->len); > } > > /* set scatter-gather table for request */ > @@ -348,7 +349,7 @@ > int i, scat_req_sz, scat_list_sz, sg_sz, buf_sz; > u8 *virt_buf; > > - scat_list_sz = (n_scat_entry - 1) * sizeof(struct hif_scatter_item); > + scat_list_sz = n_scat_entry * sizeof(struct hif_scatter_item); > scat_req_sz = sizeof(*s_req) + scat_list_sz; > > if (!virt_scat) > Index: drivers/net/wireless/ath/ath6kl/hif.h > =================================================================== > --- drivers/net/wireless/ath/ath6kl/hif.h (revision 715) > +++ drivers/net/wireless/ath/ath6kl/hif.h (revision 716) > @@ -197,9 +197,9 @@ > /* bounce buffer for upper layers to copy to/from */ > u8 *virt_dma_buf; > > - struct hif_scatter_item scat_list[1]; > + u32 scat_q_depth; > > - u32 scat_q_depth; > + struct hif_scatter_item scat_list[0]; > }; > > -- > To unsubscribe from this list: send the line "unsubscribe linux-wireless" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
Robert Deliën wrote: > This patch fixes broken hif-scatter in ath6kl SDIO. It is not my work; I have > found it as a difference between two Digi distributed Kernel source trees. No > commit comment was recorded. I am sure there are better/nicer/cleaner way, > but copying it as I found it is yet another difference less in the future. > Feel free to use it in any way you see fit. Robert told me privately that this is fixed in latest versions. The patch has been dropped.
Index: drivers/net/wireless/ath/ath6kl/sdio.c =================================================================== --- drivers/net/wireless/ath/ath6kl/sdio.c (revision 715) +++ drivers/net/wireless/ath/ath6kl/sdio.c (revision 716) @@ -222,6 +222,7 @@ struct mmc_data *data) { struct scatterlist *sg; + struct hif_scatter_item *scat_list; int i; data->blksz = HIF_MBOX_BLOCK_SIZE; @@ -240,14 +241,14 @@ sg = scat_req->sgentries; sg_init_table(sg, scat_req->scat_entries); + scat_list = &scat_req->scat_list[0]; + /* assemble SG list */ - for (i = 0; i < scat_req->scat_entries; i++, sg++) { + for (i = 0; i < scat_req->scat_entries; i++, sg++, scat_list++) { ath6kl_dbg(ATH6KL_DBG_SCATTER, "%d: addr:0x%p, len:%d\n", - i, scat_req->scat_list[i].buf, - scat_req->scat_list[i].len); + i, scat_list->buf, scat_list->len); - sg_set_buf(sg, scat_req->scat_list[i].buf, - scat_req->scat_list[i].len); + sg_set_buf(sg, scat_list->buf, scat_list->len); } /* set scatter-gather table for request */ @@ -348,7 +349,7 @@ int i, scat_req_sz, scat_list_sz, sg_sz, buf_sz; u8 *virt_buf; - scat_list_sz = (n_scat_entry - 1) * sizeof(struct hif_scatter_item); + scat_list_sz = n_scat_entry * sizeof(struct hif_scatter_item); scat_req_sz = sizeof(*s_req) + scat_list_sz; if (!virt_scat) Index: drivers/net/wireless/ath/ath6kl/hif.h =================================================================== --- drivers/net/wireless/ath/ath6kl/hif.h (revision 715) +++ drivers/net/wireless/ath/ath6kl/hif.h (revision 716) @@ -197,9 +197,9 @@ /* bounce buffer for upper layers to copy to/from */ u8 *virt_dma_buf; - struct hif_scatter_item scat_list[1]; + u32 scat_q_depth; - u32 scat_q_depth; + struct hif_scatter_item scat_list[0]; }; -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
This patch fixes broken hif-scatter in ath6kl SDIO. It is not my work; I have found it as a difference between two Digi distributed Kernel source trees. No commit comment was recorded. I am sure there are better/nicer/cleaner way, but copying it as I found it is yet another difference less in the future. Feel free to use it in any way you see fit. Signed-off-by: Robert Delien <robert@delien.nl> the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html